Predicting Sales from Online Buzz - 2

For several years now, many organisations have been actively monitoring and analysing the online debate in order to gain further insight into consumers’ wants, needs and experiences.

Increasingly, organisations are now taking online analysis a step further by using online buzz to help predict sales, market share and other outcomes, and to detect changes in competitors’ MarCom activities.

The idea is compelling in its simplicity: by listening to what stakeholders are saying about different brands, a reliable forecast can be made as to whether or not customers may prefer one brand over another; and why.

Unfortunately, however, it is not as straightforward as it may sound. Simply counting up the change in brand mentions is not good enough and may often lead to disastrous results.

A crucial element in transforming the online buzz into reliable predictions is the ability to attribute to each online ‘voice’ the correct weight; often referred to as ‘influence’. This of course is very intuitive: It counts more when somebody who is a recognised authority or has a large following on a particular topic talks about a particular brand than if somebody with no following voices his or her opinion. If a particular car brand is mentioned in the driving section of The Times it counts for more than if a competing brand is mentioned on my blog or indeed in The Sun.

Some may now be thinking that surely more sales will only be the result if a brand is mentioned in a positive context or is unreservedly recommended. However, that is not necessarily the case.
All things being equal, it is normally better for a brand to be mentioned in a positive context than in a negative one. But we have to remember that every time a brand is mentioned in a negative context there are two opposing forces at work. The first force is negative. The reader may be slightly less likely to favour the brand because of the negative context. However, because the mention of the brand increases the reader’s familiarity with the brand and brings the brand to the forefront of the reader’s mind, a positive force is at work too.

Whilst the old saying that “any PR is good PR” is not entirely true, research shows that it is, in fact, almost true. Unless the talk about a brand is either very, very negative or unanimously negative, any brand mention is likely to have an overall positive impact. The occasional negative mention actually often contributes positively to increasing outcome.

In fact, when we at Onalytica test prediction models we can see that in most cases we get a better (and very good) prediction of outcome (e.g. sales) if we leave sentiment out of the model. (However, I must stress that in certain situations the model actually improves slightly by including sentiment.)

When we started out predicting sales and other business outcomes from the analysis of online buzz, we were concerned that the data we collected online were not representative of the total debate. However, our ability to satisfactorily predict sales of goods that cannot be purchased or delivered online, i.e. cars, movie visits, and prescription drugs, based solely on analysing the online debate, has largely satisfied these concerns.

In fact, when we collect the buzz online what we get is a very large and very representative sample of the overall buzz (off line and online); when it comes to the debate about the vast majority of interesting issues and brands there is no separate online or offline debate. When there is an increase in the online debate about a brand there is also an increase in the similar debate at the pub and in the work place.

The ability to predict business outcomes from online buzz has sparked new ways of working among several of our clients.

Some now set out targets known as “influence budgets” that are similar to traditional budgets where revenue is the target except that here the target is related to how much online ‘influence’ a brand earns; on its main brand, its individual products and services, and on key marketing messages.

Using “influence budgets”, organisations can now predict more precisely whether or not they are on track to meet their actual revenue or market share targets; and if they are not on target they are in a position to take action earlier.

The actual actions initiated are often more traditional. They most often involve adjusting their MarCom activities, including (but not limited to) the total spend.

An interesting side-effect is that the process of benchmarking brands against an influence budget also gives organisations early insights into changes in competitors’ MarCom spends and their effectiveness.

An overview of predicting business outcome from online buzz would not be complete without a few comments on how some of the key elements differ across markets.

When it comes to predicting outcome from online buzz there are two main factors that differ from market to market.

The first market-dependent factor is the lag from an observed change in the online buzz to the change in business outcome. This lag may range from a week (e.g. prescription drugs) over 30-60 days (e.g. cars) to over 6 months (e.g. white goods).

The second market-dependent factor is how accurately changes in business outcome can actually be predicted. In certain segments where the goods/services are difficult or expensive to sample before purchase (e.g. cars, travel, mobile phone services, gadgets, financial services, etc.) the models work extremely well. In areas such as cheap FMCGs where the goods can easily and inexpensively be sampled, the ability to predict changes in sales may work less well, but can still compete very favourably with traditional models.

In a market place where products and services increasingly are at par, earlier access to better and more precise information is likely to become even more important.

By transforming online buzz into actionable intelligence, managers can now act earlier and on safer grounds if they are not on track to meet their business targets.


(To see an example of the realtionship between sales and online buzz see this previous post.)

Halo 3 does the job for Xbox

We have previous (1, 2) written about how Nintendo’s Wii has outperformed Sony’s PS3 and indeed all other comparable games consoles.

Until now the focus has mostly been on the battle between Wii and PS3, but in recent months Microsoft’s Xbox has been making waves.

The figure below shows the relative share of buzz about the 6 major games console brands from June 1st to 23rd of September 2007.

Notice how the buzz levels of PS3 and Wii have been relatively steady while that of PS2 and GameCube has been declining.

Microsoft’s Xbox seems to be the only one with steady growth although it seems to be relatively small.

Tracking share-of-buzz is interesting because we know from research that share-of-buzz and share-of-market converges, so if the present trend continues, Xbox is on route to a small but steady increase in market share.

However, we also know that share-of-influence and market share normally converges faster so it is normally our preferred (main) variable when trying to predict future market share.

(The difference between share-of-buzz and share-of-influence is essentially that the latter includes a weighting of each mention according to the mentioning media’s measured topical influence. When calculating share-of-buzz, all mentions essentially counts the same.)

The figure below shows the development in each brands share-of-influence for the same period.

Notice how much more dramatic the picture looks: PS3 and Xbox are losing and gaining (respectively) share-of-influence much faster than the previous graph would indicate.

This would indicate at least two things: First of all, that the coverage of Xbox more often takes place in media with above average influence on the topic of “games consoles” and similar that the coverage of PS3 is usually in media with below average influence on this topic.

Second, we can predict that the increase in Xbox market share and the decrease of PS3’s similar will be more dramatic, both in terms of magnitude and speed, than the first graph would indicate.

A third observation might be that since Xbox’s increase seems to come at the expense of PS3’s, this might indicate that PS3 and Xbox are fighting for the same audience and are considered substitutes. Those who are considering buying a Wii are more likely to be choosing between buying a Wii and not buying a console at all. If this is true then part of Wii’s success is that it is enlarging the market for games consoles with new customer segments.

Now, we also know from research that if a brand has high share-of-positive-influence its market share and share-of-influence tend to converge even faster.

However, this just makes things even worse for PS3.

The graph below shows the development in sentiment or tone-of-voice of the articles/blogs/forums where each brand appeared in the relevant period.

We can see that Wii and Xbox are at a positive sloped angle indicating that the positive mentions (dramatically) outweigh the rest. PS3, on the other hand, is represented by a flatter curve indicating that the posts on this brand are more balanced (or neutral) and thus on average less positive than the two other main brands.

So this all leads to the question of why Xbox sale is performing so well.

The answer is likely to have several reasons, but look at the figure below that plots the change in debate on Halo 3 and the change in the debate on Xbox for the analysed period.

More specifically the figure shows the change in accumulated influence on a month-by-month basis for Halo 3 and Xbox.

Notice how the lines follow the same trend. In August there was an extraordinary large change in influence for Halo 3 but while it does pull up the debate on Xbox it doesn’t do so with the full force of the change.

The figure below gives a more precise picture of exactly how closely correlated the debates on the two brands (Halo 3 and Xbox) are.

The figure plots the % change in accumulated influence for Halo 3 along the x-axis and the similar change for Xbox along the y-axis. The line is a linear trend line showing the correlation coefficient to be 0.788 which indicates a strong correlation.

Whilst we haven’t proven a causal effect here I think it is at least intuitive to assume that it is Halo 3 that mainly causes the debate on Xbox and not the other way around (although some argument can be made for that).

But if we for a second assume that the causal effect is from Halo 3 to Xbox then we can take our analysis one step further.

We can see from the graph that the elasticity of the relationship is about 0.2 (0.171) indicating that if Microsoft is successful in generating a 1% change in the debate on Halo 3 they are likely to increase the accumulated influence of Xbox by 0.2%.

Because of the strong relationship between share-of-influence and share-of-market the above could be translated into monetary value if we had estimated the relationship between share-of-influence and share-of-market for these brands. However, I don’t currently have access to good market share data for games consoles, so it will have to wait for another day.

Measuring Influence – PR Blogs – Part I

In the recent weeks a number of blog posts dealing with measuring influence have stirred up quite a debate.

Jeff Jarvis and Steve Rubel aired their thoughts on the issue.

David Brain and Jonny Bentwood from Edelman’s London office published something termed “Social Media Index” where they propose a brand new method for measuring on-line influence.

Most of my feelings after having read the article were (far better) summed up by Jennifer Mattern in her comment.

The article offers no research, no references to prior research, no logical reasoning for its claims and proposes no way of testing the model.

Can you imagine if the CEO and the Head of Research of a large investment bank published a new options pricing model without actual research and without mentioning the Black–Scholes model? Of course not. So why in PR?

Actually the very notion that influence is not related to a topic is comical. As I understand it, the article proposes that someone has the same influence regardless of the topic discussed. In other words, TechCrunch (for example) is as influential on the topic of fly fishing or jogging as it is on Silicon Valley gossip. I don’t think that’s the case. We know both from intuition and from research that influence is topical.

At best, the Social Media Index is an indication of popularity. But popularity is not influence. Those who are very popular often have good influence but you don’t need to be very popular to have a lot of influence.

When you measure popularity, all “votes” count the same. However, when measuring influence each “vote” counts with the weight of all the votes leading to the voter.

Take a look at the figure below.

Person A is clearly more popular than person B. However, those who listen to person A do not, to any large extent, go on to influence others.
Quite differently so for B. Those who listen to person B they go on to influence others, who go on to influence others, and so on. The aggregated impact (or influence) of B is in this case bigger than A.

The figure (above) indicates why models that only take the first layer into account will come up short when trying to explain the overall outcome.

In pre-Internet times measuring popularity (or “reach” or “circulation”) was a relatively good proxy for influence because those who read a particular news paper or watched a TV programme did mostly not go on to influence anyone outside their close social circle.

But because of the Internet this has all changed and the social event horizon is now defined by language rather than physical distance. Both direct and indirect influence must therefore be taken into account when trying to measure influence correctly.

Relative influence correlates better to outcome than relative popularity simply because influence takes the indirect effect into account.

So how to measure influence?

Well, it’s been done for a long time. In the academic community “citation analysis” has been used for decades to measure the influence of academic journals, articles, scholars and universities.

The principle is quite simple: You collect all references made between articles about a particular topic. The references are transformed into a large set of simultaneous equations that, when solved, provides the relative influence of each journal. Thomson Scientific is probably the leading provider for the academic community.

However, the original science for measuring influence in a linked network was developed by Wassily Leontief who devised something called the Input-Output Analysis. It was originally (and still is) used to measure how sectors of the economy directly and indirectly influenced each other.

Leontief won the 1973 Nobel Prize in Economics for this specific work.

Now, you may doubt that knowledge gained in pre-Internet times more than 60 years ago can provide any useful input to explaining how to measure online influence.

But in much the same way as engineers at NASA draw on Isaac Newton’s work, those who shape the on-line world draw on Leontief’s.

In 1965 Random House published a book by W. H. Miernyk titled “The elements of input-output analysis”. The book deals, as the title says, extensively with Leontief’s work.

This book is cited as a source in the article “Measuring the relative standing of disciplinary journals” by P. Doreian (1988).

Doreian’s article then goes on to be cited as a source by Jon M. Kleinberg in his article “Authoritative Sources in a Hyperlinked Environment” (1998).

Then, same year, an article, which you may have heard of, cites Kleinberg’s:

Lawrence Page, Sergey Brin, Rajeev Motwani, Terry Winograd. The PageRank Citation Ranking: Bringing Order to the Web.

There you have it: From Leontief to Google in 4 easy steps. (There may be shorter/more paths).

Ok, end of part 1.

In part 2: The list of the 80 or so most influential PR blogs measured using citation analysis including their relative influence. Stay tuned.

Benchmarking the influence of a blog

Another interesting example of how influence has developed over time comes from looking at the blog of Iain Dale. Mr. Dale is a prominent political commentator whose blog has gained quite a following.

The graph below shows the influence of Mr. Dale’s blog when the topic is “David Cameron”. The baseline is again the influence of the BBC (same context).

We can see that the Influence of Mr. Dale’s blog in February was roughly 8% of that of the BBC.Since then his relative influence (compared with the BBC’s) has grown steadily – until the end of June. Here it took a bit of drop.

However, it is likely that the drop was attributable to a surge in the general influence of the BBC on the topic of David Cameron.

Using the contextual influence of The Guardian as a baseline for Mr. Dale’s blog reveals that he is still doing well. (See below).Benchmarking influence aside it is actually interesting to see how much influence a blog can carry on a major topic. Mr. Dale’s influence on the debate on David Cameron currently stands at about 15% of that of the BBC and just under 60% of that of The Guardian.

Benchmarking Influence

Measuring how the influence of a stakeholder develops over time is a topic of great interest to many of our clients.

Some clients are interested in understanding the effects of a marketing/PR effort. Others may be interested in identifying new rising stars among the influencers.

However, benchmarking influence over time can sometimes be a bit tricky. Because influence is relative we have to benchmark the influence of a stakeholder relative to one or more other stakeholders.

Using InfluenceMonitor I took a look at how the BBC’s influence on the topic of climate change has developed.

As a baseline I used the influence of IPCC, the Intergovernmental Panel on Climate Change.
Figure 1 (below) shows the relative influence of the BBC as a percentage of that of the IPPC since March 1st this year.

The BBC used to be more influential than IPCC on this issue, but around the end of April its influence dropped below that of IPCC. We can see this as the relative influence drops below 100%. Notice, however, the spike in BBC’s influence at the very end of the graph. The increase in BBC’s influence is a good illustration of how a heavily discussed article can dramatically increase influence.

On the 10th of July BBC published an article about a new study that (apparently) concludes that there is no link between the cosmic rays from the sun and global warming. (See partial screen shot below).

As can be seen from the graph below, this article caused a spike in the number of references and inbound links to the BBC, in the context of climate change, over the following days.

Buzz, Influence and Impact

When it comes to analysing online/social media, “buzz” has long been the buzzword.

“Buzz” is of course the industry slang for the online “talk” about a particular topic or brand.

There are several free online services that will show how much buzz there is about a topic or a brand. Some will even show how big a percentage of the buzz on any given day is about a particular topic.

These services are great for getting a general impression the buzz levels.

However, when it comes to making conclusions based on buzz levels, one has to be careful relying on “raw buzz”.

One of the problems with measuring the levels of raw buzz are that raw buzz treats all voices with equal weight: 1 mention in a random MySpace blog and 1 mention in The NY Times online edition both count as 1 mention and contribute equally to the raw buzz level about a topic.

However, the impact of a mention in the NY Times is likely to be greater than being mentioned on a random MySpace blog. Why? Because NY Times carries more influence on most (if not all) topics than your average MySpace blog.

So in order to make more sense of buzz one needs to factor in the weight (influence) of the “buzzer”.

The figure below shows the relative amounts of online buzz about 3 pharmaceutical brands got in the context of “Blood Pressure” in the 14 days prior to the 8th of March 2007.

So according to the chart, Pfizer was mentioned 2.7 times more often than Merck in the context of “Blood Pressure” during those 14 days.

During the period Merck and Pfizer was discussed in a variety of online media. Some discussed both brands; others just one of them.

And because they appeared in media with different influence on the topic of “blood pressure” the figure showing the distribution of Raw Buzz doesn’t give the correct picture about how much impact the brands made.

During the 14 days Pfizer appeared in the context of “blood pressure” in the NY Times, International Herald Tribune and Medgadget.com as well as a list of other online media. All of these 3 sites carry more influence on the topic of “blood pressure” than Businessweek.com – the most influential onlinepublication Merck was mentioned in during that period (1).

The figure below shows the relative impact of the 3 brands during the period. Because of the greater influence of the media Pfizer appeared in, they actually made an impact that was 4.6 times that of Merck for the period – not 2.7 as the raw buzz would suggest.

The two figures below shows the same distribution of raw buzz and impact – now monthly over 90 days.

Notice how Pfizer had a 73% share of Raw Buzz in December vs. Merck’s 26% - while the actual impact distribution was 46% vs. 53%. This is a good example of how the interpretation of the Raw Buzz could have led us to a wrong conclusion about the impact of the online media coverage.

Influence not only plays an important role as a way of converting buzz to impact. It is also crucial to get a correct impression of the sentiment surrounding a topic or brand.

Although some may argue that the next example deals with neither a topic nor a brand it illustrates the point.

The figure below shows the number of online blog post about British politician David Cameron for the last 30 days.

The orange graph shows the aggregated sentiment (Net Promoters Index) of the posts. Notice there is a steep dive on the 11th of February. On that day a number of news reports suggested that Mr. Cameron had smoked cannabis during his younger years.

This news, combined with Mr. Cameron’s media handling, prompted an increase in negative posts and hence the drop in sentiment.

However, looking at the graph one could easily come to the conclusion that the sentiment had now stabilised at a lower level. This would not be too bad for Mr. Cameron as the direction of the sentiment is normally more important than the absolute level.

However, if we take a look at the next graph, where the sentiment is adjusted with the relative influence of the “speaker” the picture is quite different and more positive for Mr. Cameron.

It shows that although he took a hit on the 11th of February he has not only recovered but is basically continuing on the very positive trend from before.

When comparing the two graphs we can conclude that those who are more influential on the topic of “David Cameron” are more positive than those who are less influential.

The take for Mr. Cameron’s team could be that the issue from February 11th is now largely over and Mr. Cameron’s PR people should consider leaving the matter alone. This might not be the conclusion they would arrive at had they just relied on Raw Sentiment rather than Impact Adjusted Sentiment.

(1) The relative influence of the 4 sites, on the topic of “blood pressure” was measured to 4.02, 1.62, 1.42 and 1.20 respectively – measured over 120 days

For more on how to measure influence, take a look at one of my previous posts here

All pictures are from the Onalytica InfluenceMonitor and © Onalytica

Who are influential on the debate on innovation in the UK?

At Onalytica we recently concluded an analysis of who are influential in the public debate on “innovation” in the UK.

A few results are listed below.

Please contact us if you have an interest in the full report.

Table 1 (below) shows the 25 most influential stakeholders of the topic of “innovation in the UK”


Table 1 (above) also shows the popularity, influence and “relative influence” of each stakeholder.

The DTI (Department of Trade and Industry) is both the most popular and the most influential stakeholder of the issue and both popularity and influence have been indexed to DTI=100.

In general those who have funds to spend on supporting innovation as well as those who are attached to the topic for legal reasons (like UK Patent Office) top the list of influencers.

In short terms, popularity is how many thinks a stakeholder is relevant. Influence is who thinks you are relevant. When measuring popularity each “vote” counts equal, but when measuring influence a stakeholder “votes” with the weight given to her by votes from other stakeholders.

By estimating the typical relationship between influence and popularity on this issue, we have calculated the influence a stakeholder should be expected to have, given their popularity. We then compared this with their actual influence. The result is shown in the column “Rel. Influence”.

So when BBC has Rel. Influence of 88% it means that the BBC has 88% of the actual influence on this issue that their popularity could lead us to believe. In other words they are not quite as influential as they are popular.

CORDIS, MIT and IBM are not UK entities but they are still highly influential in the debate as they are often cited in the context of “Innovation in the UK”.

Notice how Blackwell, a publisher of scientific journals, has substantially more influence than their popularity should warrant. Not entirely counter-intuitive when thinking about how influence is measured. Perhaps those who cite one of Blackwell’s journals in this context, on average, have more influence than those who cite the BBC?

Table 2 (below) shows the 25 most popular stakeholders of the issue with several stakeholders from Table 1 reappearing but in new positions.

Table 3 (below) shows the top 25 “over-influencers”.

These are the stakeholders who have substantially more influence than their popularity should warrant.

Notice that most of them have pretty low popularity. They don’t receive many citations in the analysed context but when they do get cited, they often get cited by very influential stakeholders or by influential stakeholders who cite very few other stakeholders.

Notice the absence of popular media brands.

Table 4 (below) shows the 25 most “under-influential” stakeholders.

These are stakeholders who are very popular on the issue but at the same time have less influence than their popularity should warrant.

Notice that the combined influence of the stakeholders in Table 4 is still substantially bigger than that of the stakeholders in Table 3.

Stakeholders who are generally well-known and very popular on a particular issue tend to become more “under-influential” the bigger they get. It doesn’t mean they get less influential. It simply means they may not be quite as influential as their fantastic popularity on the issue could suggest.

One can also view it as a result of successful brand building. Those stakeholders who have become household brands are more at the forefront of people’s mind, and may therefore more often get cited to substantiate an argument.

One interesting observation we made during this analysis was that the correlation between popularity and influence as stronger and more linear than usual. In our experience this is observed more often with ambiguous topic. (“Innovation” is not a super precise topic. Depending on the context, it can mean different things to different people).

How we did (short version).
We downloaded all documents (that we could identify; including web pages and other documents) freely available online mentioning “innovation” in a UK context. From these documents we extracted all contextual citations (where the author cities someone else). These citations were then counted (to find popularity) and turned to a system of equations to find the influence-equilibrium; fully taking both indirect influence and attention span into account.

© Onalytica, 2007

Attending the Maximising Impact conference

Next Wednesday (22nd November) I am attending the Maximising Impact conference here in London.

The conference focuses on measuring the impact of Corporate Social Responsibility (CSR) initiatives in general and educational CSR initiatives in particular.

I have been invited because Onalytica is supplying some of the analysis work and cases used at the conference.

Measuring the reputational impact of CSR initiatives is an issue of increasing interest for several of our clients.

Traditional measurements of CSR initiatives have focused on benchmarking issues like the scale of reductions in CO2 emission, waste and general environmental impact.

However, many (if not most) organisations engaged in CSR initiatives are increasingly focused on the value CSR can generate their corporate reputation and/or authority on particular issues.

Because of this, an increasing number of organisations ask us to measure the influence of stakeholders of specific issues that are closely related to their CSR activities.

This way, organisations can not only benchmark how their own influence and authority on a particular issue develops. They are also better positioned to identify organisations they can team up with who already have influence on the issue.

By understanding how their own influence and authority on CSR issues develops organisations can better ensure they get value for money from their CSR investments.

This insight also enables organisations to better value and compare different partnership opportunities.

Please drop me a mail if you are attending the conference and would like to meet up. (flemmingm [-at-] Onalytica.com)

Previous related posts
Who are influential authorities on child obesity? (also measures the influence of Jamie Oliver)

Who are influential authorities on CSR in the United Kingdom?

Edelman, influence and intelligent design.

Edelman, a PR company, and Technorati, a blog-search-engine, has released a list of "the most influential bloggers".

According to Edelman’s own release the methodology has been to count the number of links each blog receives from other bloggers.

If that is a correct description of their methodology it can be deduced that their analysis rests on two central conjectures that are both wrong.

1st wrong conjecture: Influence is defined using one factor only: The number of endorsements (links) an actor receives.

2nd wrong conjecture: Influence is independent of issue.

In a social context influence (often in the literature referred to as “prestige”) is normally defined using at least two factors: The number of endorsements an actor receives and the prestige of each actor awarding the endorsement. Lately (in the last 30 or so years) a third factor is usually also included: The number of actors each actor is awarding his or her prestige to (linking to) (i.e. how thin they spread their total endorsement).

Popularity, on the other hand, is usually defined using one factor: The number of endorsements an actor receives.

The Edelman/Technorati study doesn't attribute different weights to the links/endorsements by different bloggers. One could say that they regard a link from any two bloggers as contributing equally to the influence of the blogger linked to.

This (1st wrong conjecture) is intuitively wrong as well as logically flawed.

Intuitively wrong because what they are saying is that it would give a blogger equal influence to receive the endorsement of the most credible and well respected blogger in the world as an endorsement made by someone who hasn't got a clue and that nobody has ever heard of. Intuitively it doesn't make sense.

The conjecture is also logically flawed. The very foundation for their methodology is that all bloggers are equally influential (as all links count equally). And what do they produce: A ranked list of bloggers stating that some are more influential than others. But hey - what do you actually mean guys: Should a link from the number 1 on your list count the same as a link from the last guy on the same list? It does when you start counting but it doesn't when you're done. Clearly there was a change of mind (and logic) somewhere along the way..

As for the 2nd wrong conjecture it is even more illogical than the first. (This is the one that says that influence is universal).

Are you really saying that, if say David Beckham is influential on football he is also influential on the developments in British politics and on say, wine?

Influence is issue-based because the endorsements/links are related to the issue discussed. If not, bloggers would link to the same blogs in any context and this is obviously absurd.

Reading that influence is independent of issue reminds me of the time (Feb 17th, 2006) I was researching a story on bird flu and searched Technorati to see who they regarded as the most authoritative blog on this issue. The result: Engadget – a blog on gadgets.

The whole thing leaves me with a number of questions I don't seem to be able to answer:

Why don't they just call it "popularity"? Then they would be totally right. They could have said "We have made a list of those bloggers who are most popular with other bloggers". That could have been a respectable piece of research and an interesting list. as those who are popular usually also have some influence (although you don't need to be popular to have influence).

Why do they call it "influence" when it's clearly not?

Thanks to Antony Mayfield for pointing me to this.

Is Jamie Oliver influential – or is he just popular?

Jamie Oliver and child obesity
At Onalytica we recently completed an analysis of who is influential in the debate on child obesity in the United Kingdom.

One of the objectives of the analysis was to determine if those companies who have corporate social responsibility (CSR) initiatives relating to child obesity are getting value for money: Are they being considered relevant authorities in the debate on child obesity and do they have influence over and above what one might expect given their popularity?

One of the more intriguing results of the analysis is that it quantifies the influence and popularity of Jamie Oliver, a British celebrity-chef, on this issue.

Combining Mr. Oliver’s celebrity status with the noble cause of improving the diet of British school children has been a hit with popular media. So while there is general consensus that he has succeeded in raising the public awareness on the issue, the magnitude of his actual influence is a matter of often heated debate.

Popularity vs. influence
While it is easy to establish the popularity of a person, organisation or institution (hereafter collectively referred to as “stakeholders”), it is more complex to determine the influence.

To measure a stakeholder’s popularity on a particular issue all one has to do is to count the number of other stakeholders who references said stakeholder in the appropriate context. In doing so we regard all stakeholders as equally important; a “vote” carries the same weight no matter who gives it.

Not so when it comes to influence. Here the weight of a stakeholder’s “vote” on how influential another stakeholder is, is determined by the “voting” stakeholder’s own influence, whose influence again is determined by the influence of those who vote for him or her and so on, and so on..

So to put it short; popularity is about how many listens to you whereas influence is more about who listens to you.

The results
Table 1 shows the 50 most influential stakeholders of the debate on child obesity in the United Kingdom.

The NHS is in a league of their own followed by The World Health Organisation (WHO), Department of Health, The European Union and the Food Standards Agency (FSA).

Mr. Oliver’s relative influence is 4.74, roughly equivalent to the influence Sport England and British Nutrition Foundation. His influence on this issue is not only greater than of the Guardian, Department for Culture, Media and Sports, 10 Downing Street but also somewhat greater than the influence of Tesco and Coca-Cola Company.

Because the scale in Table 1 is linear we can see that Mr. Oliver’s influence is roughly half of the BBC’s influence on this issue.

Table 2 (below), on the other hand, shows the 50 most popular stakeholders of this issue. (The popularity scale is relative and has been normalised to match the influence scale, but beware, it’s not possible to compare the numbers directly between the Tables 1 and 2).

Table 2 shows that NHS is also the most popular stakeholder of the issue. In fact many of those who are at the top of Table 1 have good positions in Table 2. However, if we compare Tables 1 and 2 closely we will see that the order and position of many of the stakeholders who are listed in both tables are somewhat different.

The relationship between popularity and influence is not linear and can differ quite a bit depending on the issue analysed. But it is possible to approximate a formula that quite elegantly transforms popularity (of a particular issue) into influence.

This enables us to calculate what kind of influence we should expect a stakeholder to have given their popularity. This further enables us to calculate who is more influential than their popularity should warrant and vice versa.

Table 3 (below) shows the details.

In Table 3 “Popularity” refers to the relative popularity of the stakeholder as listed in Table 2. “Expected Influence” refers to the influence that we should “expect” from the given popularity. “Actual influence” refers to the correctly measured influence as listed in Table 1. Finally “Under- Influence” refers to how much less influence the stakeholder has compared to what we should expect.

From Table 3 we can see that BBC’s under-influence is 28%. So the BBC is then quite a bit more popular than they are influential when it comes to child obesity. This is not surprising when we examine the reasons behind why someone may be under-influential.

A stakeholder is typically under-influential for at least one of the following two reasons: Either those who are influenced by the stakeholder with under-influence have less than average influence, and/or those who are influenced by the stakeholder with under-influence are influenced by many other stakeholders.

Another way to look at under-influence is as a measure of “over-popularity”. It is equally fair to say that the BBC is 28% more popular on this issue than they are influential. It is hardly surprising that BBC is more popular than influential as they have a huge public service obligation that requires them to be popular.

Table 3 basically shows who has been successful in establishing themselves as popular stakeholders on the issue of child obesity. The keyword here is “successful” because appearing in Table 3 means that you get more coverage than your influence should warrant.

We can see that Mr. Oliver’s under-influence/over-popularity is 9%. This may come as a surprise to those who thought he was only being listened to by the celebrity media and that he had little or no real impact. If that was the case his under-influence would have been far greater.
So to summarise, Mr. Oliver has real influence and much more so than many other high profile stakeholders. Also, Mr. Oliver is being listened to by stakeholders with real influence on this issue, not just the celebrity magazines – who may have a lot of influence on fashion and glamour but limited influence on the issue of child obesity.

Last week I mentioned this upcoming analysis to a leader of some of NHS’s mental health initiatives. When I mentioned that Mr. Oliver does have real influence and impact on this issue she replied “it’s what I have always said – we need a Jamie Oliver to focus on mental health”. It looks like she was right.

Table 4 (below): List of abbreviations

(C) Onalytica 2006

Onalytica on "For Immediate Release" podcast

The well-known podcast “For Immediate Release” by Neville Hobson and Shel Holz has a special edition about the recent Onalytica study on popularity vs. influence in relation to Blog Marketing.

Listen to it directly or download it here

Who are the most influential authorities on “blog marketing”?

(A more comprehensive version of the report is available for download here (PDF, 10 Mb))

In a previous analysis we analysed who are influential authorities on “business blogging”.

Since then there has been great interest in both the results of the analysis and the methodology used.

We decided to do a new analysis on another topic of interest to many bloggers but this time also use some time to make a point of illustrating influence versus popularity.

In the new analysis we set out to identify the most influential authorities on the topic of “blog marketing” and compare this list to the list of those who are the most popular stakeholders of that topic.

The difference between influence and popularity
The difference between influence and popularity can be highlighted by looking at the parameters taken into account when measuring the two.

When you want to measure the popularity of a stakeholder, you count the number of other stakeholders who refer to the first stakeholder in the context being analysed. So the only variable taken into account is “number of referrals”.

Analogous this can be used to measure the link-popularity of a particular website by counting the number of inbound linkers.

However, when measuring influence we take one more variable into account: The influence of the endorser (linker).

The influence of academic journals and universities have been measured this way for more than 30 years, but it also intuitively makes sense: It typically means more to any of us to receive the endorsement of someone we regard as an authority in the field than from someone we know hasn’t got a clue.

So in order to establish popularity all we have to do is count. But to establish influence we have to turn the references (or links or endorsements or citations) into a system of simultaneous equations and solve them to find an equilibrium state. (Please see the previous analysis on “business blogging” for a more in-depth discussion on these issues).

But aren’t those who are popular in relation to a topic also influential on that topic?
The short answer is “Yes, they often are”.

The little longer answer is “Yes, they often are, but you don’t have to be popular to be influential”.

In that answer lies one of the main reasons why it is interesting to identify who are influential on a topic: If someone is influential, but not really popular, then we may not know about them and we are therefore not able to take them into account.

In a public relations context there is a more interesting side to this thinking as well. Someone who is less popular may be easier to influence than someone who is popular. And if their influence is roughly the same, it may be more cost-effective to influence those who are not so popular.

The findings of our analysis may underline my point here.

Results
Table 1 shows the 20 most influential authorities on the topic “blog marketing”. The table also shows their popularity. (Both measures are relative and linear.)

The table shows that New York Times is the most influential authority on “blog marketing”. Interestingly enough it is also the most popular. (To be valid the reference must occur in the relevant context. In this case, a reference to New York Times must happen in a “blog marketing” context to count.)

Table 1 contains many of the usual suspects: Seth Godin is there, so is Steve Rubel from Micropersuasion (now Edelman PR), BusinessWeek, Fast Company and a handful of other well known names.

But there are also some surprises, or “not so popular” names like Hyku, Next Level Biz Tips, All Business and Twist Image.

Figure 1 shows the whole thing in graph format.

The difference between the influence bar and the popularity bar for each stakeholder is a measure for how much they “punch above their weight” so to speak. How much more influential they are than their popularity should lead us to believe.

The opposite is true for Micropersuasion, BusinessWeek, WebPro News, Problogger, Buzzmachine, and Adrants. They are all somewhat less influential than their popularity would suggest.

Table 2 and Figure 2 shows the similar information for the 20 most popular stakeholders of the topic analysed.

Table/Figure 2 have several names from Table/Figure 1; but also some new ones. Big media brands like BBC and CNN are showing. They are popular and they do have influence – but not as much influence as their popularity suggests.

Actually, among the 20 most popular stakeholders only Search Engine Watch and Top Rank Results have more influence than their popularity would suggest and “punch above their weight”, so to speak.

The results demonstrate that sheer popularity can give influence, but also that popularity is not a requirement for influence.

Digg this story

Personal Video Recorders in the UK - Influential Authorities

In the previous case studies in this blog we have mostly analysed influence on issues of a political nature or of general public interest.

As part of a project to demonstrate how influence measurements can be used in relation to consumer goods we decided to analyse who are influential authorities on Personal Video Recorders (aka. PVR) in the United Kingdom.

Figure 1 (below) shows the list of influencers and their relative influence. (For an in-depth description of our methodology please read this previous analysis)


If you wanted to spread the gospel about products or services in the PVR space you could try to go directly to Guardian, Business Week or other well known media brands who write about the PVR topic. But knowing how these publications gets flooded with similar pitches the chance of getting noticed could be small.

A more effective strategy may be to turn your attention to some of the smaller more specialised publications like C-Net, Silcon or Wireless Digest Blog who carry a substantial influence on the topic in the UK – not at least because they are quoted by large media like Guardian.

The Hobson & Holtz Report discuss authority vs. popularity

The well known podcast “For Immediate Release” (#113) by Neville Hobson and Shel Holz uses the Onalytica report about Business Blogging as a starting point for a discussion on authority vs. popularity.

They have several interesting perspectives on the issue and I urge anyone with an interest in the topic to listen to their show.

Aside from the above discussion the podcast as a whole is a comprehensive update on what is happening in the world of blogging, PR and new media. It’s easy to understand why their show has such a huge following.

Bird Flu - Who are the most influential authorities?

The fear of a possible bird flu pandemic is causing concern to governments and health organisations around the world.

We decided to measure who the world relies on when it comes to information about bird flu.

To measure influence we use a scientifically recognised methodology called citation analysis. (See a more detailed description after the results section)

Measuring influence this way assumes that when a person mentions another person in a particular context then it is because the former person thinks the latter is relevant to the context.

And, since the former thinks the latter is relevant to the context the latter has some influence on the former.

The actual amount of influence is initially irrelevant. It’s whether or not there is influence that matters.

The practicalities of measuring influence this way are to first find everything that has been written about an issue. We achieved this by automatically downloading everything we could find that was freely available on the internet that contained any of the following words or phrases:

“bird flu”
“avian influenza”
”avian flu”
”H5N1”

We then identified who was referencing whom in this context. (A reference can be a textual reference or a hyperlink).

These references were then used as equations in a massive simultaneous equation system that produces the influence of each stakeholder as a result.


Table 1 (above) shows the top 50 influencers on the topic of “Bird Flu”. The Issue Influence Index™ is a linear influence scale ranging from 1 and upwards. An index value of 10 thus means “twice the influence” as index value of 5.

Notice the substantial influence of the top three organisations, WHO, OIE and FAO.

The United Nations in Vietnam most likely gets its influence because several early cases of bird flu and fatalities amongst the human population in Vietnam.

Roche is most likely on the list because they are the principal supplier of the Tamiflu, an anti-influenza drug believed to have some effect on the current virus, H5N1.
(It is clear that the results favours English language media as search parameters, with the exception of “H5N1” are in English.)



Figure 1 (above) shows how the top 50 influencers reference each other. Please bear in mind that all identified stakeholders outside the top 50 have been removed from the network picture.

You can download a more comprehensive version of the results and the methodology here (0.7 MB pdf)

Who are the most influential authorities on “Business Blogging”?

While many corporations initially saw blogging as a potential treat to their reputation an increasing number have now started to use blogging as a new way of engaging with their stakeholders.

The rush for corporations to create and execute effective blog strategies have become a healthy business for a large number of bloggers who have demonstrated their ability to gain attention via their own blogs.

To understand who is influential when it comes to “business blogging” we decided to measure it.

You can download the entire report here (pdf 0.1 MB)

What is influence and how is it measured?
In the internet community in general and in the blogsphere in particular there has been a growing understanding that search engines such as Technorati measure influence. They don’t. They measure popularity which is something totally different.

When Technorati ranks blogs they count the number of link sources pointing to a blog. So a blog that has 10 inbound links has higher rank than one that has 5 inbound links. So far so good. The blog with inbound links from 10 different sources is clearly more popular than the one with 5 link sources.

However, when they use this measure of popularity as “authority” they are stretching it too far.

David Letterman may be popular when it comes to the topic of US national politics, but few would call him an authority on the topic.

(In Onalytica we don’t use the word “authority”, but are more focused on “influence”, but in this context, “influence”, “authority” and “relevance” are closely related.)

Before moving on to explaining how influence is really measured I want to give a few examples why popularity is not a measure of influence, authority or relevance.

Imagine two websites A and B. They both write about the same topic – say US politics. Website A has inbound links from 2 sources and B from 10 sources.

B is clearly more popular than A, but is B also more influential?

We can’t know from the data above.

Now imagine that website B is a blog written by a random 11-year old child as an assignment and those 10 who link to B are his or her class mates and the teacher of the assignment.

Imagine further that those who link to A are The New York Times and The Washington Post.

Intuitively website A is now the more influential. Why? Because it is deemed relevant by websites who are themselves influential on US Politics.

The above example touches on the first major reason why popularity and influence is not the same: All websites are not equal.

As in real life, your parents and family members may think you’re the greatest authority on something, but unless they themselves have authority on the issue, it doesn’t really make you an authority.

If popularity was a measure of authority (or influence or relevance) then two websites, where the first has inbound links from the 500 most trafficked websites would have the same authority as one who only had inbound links from the 500 least trafficked websites. This is clearly not the case.

The last major complaint about using the number of inbound links as a measure of anything but raw popularity is that the number is not related to the context.

If you search Technorati for “bird flu” and only want to see posts with high authority you get a page full of links to large news media. Google News came up on top when I did it.

Nobody believes that Google News is an authority on “bird flu”. They are a news aggregator.

The reason why they appear as the number 1 result is because they have an enormous number of inbound link sources.

But are those who link to Google News linking to it in the context of “bird flu”? Some may be, but I think it’s safe to assume that people link to Google News from all sorts of contexts.

As a consequence of mixing up non-contextual popularity with authority on issues, the major news media appear to be the biggest authorities on anything.

In fact, because Engadget, a hugely popular blog about “gadgets” has so many inbound link sources it appeared twice on the first page of the search results for “bird flu” in Technorati (filtered for maximum authority). I’m sure it will come as a surprise to even the writers of Engadget that they are one of the world’s foremost authorities on bird flu.

So how is influence/authority/relevance then measured?

The answer is that you have to take the indirect influence/authority/relevance into account.

In the academic community there is no real debate about how to measure influence.

For more than 3 decades academics have used something called “citation analysis” to measure the influence of academic journals, researchers and universities.

In academic articles, writers cite the works of other academics. They do that for several reasons, but mainly because they believe that those they cite are relevant to the context. They point to other publications that are relevant to their arguments and to the context. In doing so they reveal which other publications have influenced them.

In citation analysis these citations from one journal to another are regarded as links.

These links are extracted and transformed into a huge system of equations. When solved the result is a relative measure of influence.

This way of measuring influence was developed by Russian born American Wassily Leontief. He developed something called Input/output analysis to measure how sectors of the economy influence each other.

If you’re one of those who find joy in understanding how complex matrix-mathematics can give simple answers to complicated questions, then you will love his work.

The Nobel Committee did. They awarded him the 1973 Nobel price in Economics for developing input/output-analysis and thereby solving the illusive problems of “circular influence”.

How to measure the influence on an issue
To make influence measurements operational (and relevant) they have to be tied to a context (or brand, company, etc.).

This is achieved by extracting only those references that are made in the relevant context of focus.

When calculating influence we make the basic assumption that a person references another person if the former thinks the latter is relevant to the context.

We assume this logic is systematic, meaning that this is a general reason for referencing others in a particular context.

It doesn’t matter that people get referenced for other reasons (perfunctory reasons, reasons from limited knowledge, etc) as long as the same people (or websites, stakeholders, entities, etc) do not get systematically referenced when they are not believed be relevant.

The practical steps to gathering the data and measuring influence on an issue are:

First we define a search criterion. This can be simple or a set of rules. Simple ones typical give best results.

In this case our search criterion was to look for documents (web pages, blogs, pdf files, documents) that either contained the phrase “business blogging” or “business blog”.

Using our own issue focused internet crawlers any document matching the issue was downloaded and analysed for references. (A reference can be a hyperlink or a textual citation. A textual reference to “The White House” would be treated equal to a link to www.whitehouse.gov)

The references are extracted from the documents and after some semi-manual consolidation and statistical filtering they are transformed into a massive system of simultaneous equations, consistent with Leontief’s directions.

Once the equations are solved we have, viola, the relative influence of each stakeholder of the issue. We term this metric Issue Influence Index™.

The Issue Influence Index™ is a relative and linear measure of influence. It ranges from 1, which can be interpreted as “very little influence, but still more than no influence” and upwards.

An organisation with an index of 4 has twice the influence of someone with an index of 2.

The Results

Table 1 shows the top 25 influencers on the topic of “business blogging”.

Many of the names will be familiar to people with an interest in the topic.

Comments on some of the findings:

Corante
Corante is actually a group of well known bloggers operating more or less under a common brand.

Business Week and Forbes
They have published some of the most widely cited articles on “business blogging” and their attention to the topic signals to many an acceptance of “business blogging” in the business media.

Neville Hobson
Neville Hobson is the most influential appearing under his own name. To some more known as the publisher of the popular podcast “for immediate release”.

BBC
Blogging is related to news and therefore it’s not surprising to see the world’s largest news organisation on the list.

Micro Persuasion
Blog run by Steve Rubel, a famous blogger with a foucs on PR. During the week this study was made it was announced that Steve Rubel is joining Edelman, a large PR company with big customers like WalMart. It’s not difficult to understand why Edelman poached Mr. Rubel. They are buying a lot of influence on a topic of great interest to their clients.

Seth Godin
I have yet to meet someone who works with the Internet who doesn’t know who he is. Well known author and often seen as the inventor of “permission based marketing”

Ross Mayfield
Founder of the company SocialText.

GM Blogs
The entry to General Motors’ blogsphere. Often referenced as an example of how a large and relatively conservative organisation has embraced business blogging.

Report compiled during the period February 12-17 2006 by Flemming Madsen, Onalytica.

You may quote from this report when clearly referencing Onalytica as the source.

© Onalytica Ltd 2006, all rights reserved.

Who has influence on the debate on Stem Cells in Scandinavia (Denmark, Norway and Sweden)?

We recently published a report on who publish the debate on Stem Cells in the Scandinavian countries (Denmark, Norway and Sweden).

Table 1 shows the top influencers on the topic of “Stem Cells” in the Scandinavia.

The organisations in table 1 all have an Issue Influence Index™ on this topic of 1.8 or more.
Issue Influence Index™ is a generic measure of influence.

It measures both direct and indirect influence and is calculated like a citation index used to calculate influence of academic journals.

The scale is linear, ranging from 1 (one ) and upwards. An index of 1 can be interpreted as “no particular influence”. A stakeholder with an index of 4 can be interpreted as having twice the influence as someone who has an index of 2.

The table shows that the most influential organisation is the US Department of Health and Human Services’ National Institutes of Health (NIH). NIH is an index of health resources and may not provide any data of their own.

NIH is cited by many of the other stakeholders in the relevant context (stem cells) and from a strictly mathematical point of view they are the most influential. However, one can argue that as they do not provide their own content they mainly convey the influence of others. That said the editorial process of NIH whereby they include some resources and exclude others (that they don’t deem relevant) is in fact a way of providing content of their own. So as part of their editorial process they do exercise their own influence.

The second most influential organisation is the BBC. The BBC is the world’s largest news organisation. The annual budget of the BBC is quite precisely 10 times bigger than that of University of Copenhagen. When an issue is on the public agenda – such as stem cells has been – the BBC usually plays a role in shaping the public opinion; especially in the UK, the north western part of the EU and Scandinavia.

It is interesting to observe the absence of Danish organisations among the top influencers. Apart from BioMed Community, a community representing the Bio- & Medical technology competences in Aalborg, Denmark and BioTIK a portal focusing on bio-ethical questions, the most influential Danish player is Retsinformation a state sponsored information system on legal issues. This could indicate that legal red tape is relatively more of a factor in Denmark than in the two other countries.

The most influential representatives from Sweden and Norway (University of Oslo, University of Lund, Karolinska Institutet and Uppsala University) are largely those one would expect to have a prominent position, the most influential research institutions from Denmark are University of Southern Denmark and University of Aalborg; two of the newest universities in Denmark.

Figure 1 (below) shows how organisations from Table 1 (above) reference each other.

The direction of the arrow shows the reference. The influence is consequently the other way.

The size of the dot representing each organisation is proportional to their total influence.

(Unfortunately the program generating the picture does not support Unicode characters which are used to represent Scandinavian letters.)

An extended resume of the report can be downloaded here (pdf, 2 MB)

Who influence the debate on Incontinence?

We recently published a report about who influence the debate on Incontinence in the UK.
The top 25 influencers are listed in the table below. (Please see the full report for the complete list)

Table 1 (above) lists the top 25 influencers on the topic "incontinence" in the United Kingdom at the time of the analysis (Jan 2006)

Table 1 also lists the Issue Influence Index™ for each organisation.

The Issue Influence Index™ is a linear influence index that measures both direct and indirect influence using "citation analysis". (See the section on influence and methodology for further information).

The fact that the index is linear means that an organisation with an index value of 10 has twice the influence as an organisation with an index value of 5.

Figure 1 (above) shows how the top 25 influencers from Table 1 influence each other. The size of the dot representing the organisation shows their relative influence (according to Table 1)

The arrow shows the direction of the citation. The influence consequently flows the other way.

Notice how the BBC draws information from both most influencers.

Notice also how the health care sector and academic institutions seem to cluster on the left side of the network where as NGOs cluster on the right side.

© Onalytica 2005

The information in this article may be quoted when Onalytica is credited as the source. If quoted online, please add a direct web link to www.onalytica.com

Issue Influence Index™ is a trademark of Onalytica Ltd.

A resume of the report can be downloaded here

Who has influence on bills in Parliament?

Who has influence on bills in Parliament?

Onalytica today launched a new website that monitors who influence selected bills as they pass thought the United Kingdom Parliament.

We expect to track a handful of bills, updating the influence metrics every month or so.

As we are not political commentators we will not be publishing much commentary along with the influence metrics. We will leave that to others.

The first bill to be analysed was the Charities Bill.

The results and the new website can be found at www.BillInfluence.com

Measuring blogging influence on brands

Should brands be concerned about the influence of bloggers?

Together with Market Sentinel and Immediate Future we examine the “famous” case of blogger Jeff Jarvis problems with Dell.

This white paper sets out to answer questions we are often asked by customers:

• How important are online commentators to my business?
• How important are bloggers in particular?
• Can their influence be measured?
• If so how?
• Are bloggers really a unique threat to corporate reputation, as Forbes suggests, or a huge opportunity as Business Week implies?
• And if so why?

Download the White Paper here (7 MB, pdf).

Who influence Corporate Social Responsibility in the UK?

According to Wikipedia, Corporate Social Responsibility (CSR) “is an expression used to describe what some see as a company’s obligation to be sensitive to the needs of all of its stakeholders in its business operations.”

In order to identify the most important influencers on CRS in the United Kingdom an Onalytica Stakeholder Analysis was conducted.

(You can download a more comprehensive summary with network maps here - 250k pdf)

Influence
You exert influence on people when they listen to what you say or read what you have said – and vice versa.

We form our opinion about all sorts of matters using information that has been supplied to us by other parties. Is the weather getting warmer? Is the ice at the North Pole melting? We may have an opinion on this but have we actually measured it ourselves? Most of us haven’t – we rely on news media we trust to form our opinion. But when we take in information from other sources, they influence us.

When calculating influence we gather information available in the public domain and analyse it to find out who is referencing whom when it comes to the issue we are analysing.

The way we calculate influence is equivalent to the way influence of academic journals and universities are calculates: using citation analysis.

At the heart of this type of influence measurement is a simple, but central conjecture:

Person X has influence on Person Y regarding a particular issue if Person Y is dependent on Person X for information about the issue.

(“Person” can mean organisation, website, person, etc, according to the context.)

In academic citation analysis this is put into practice by a slight rephrasing:

Person X has influence on Person Y regarding the issues covered in the academic paper, if Person Y cites person X as a reference in the paper.

When measuring “issue influence” Onalytica uses the very same principle.

Based on the principles above systems of equations can be formed and influence calculated.

One of the great advantages of this kind of influence measurement is that it takes indirect influence into account.

Methodology
To construct an Onalytica Stakeholder Analysis (OSA) a focus issue needs to be defined. In this case “Corporate Social Responsibility” was chosen. (The issue can be a simple set of words or a more complex set texts and rules.)

Onalytica’s proprietary issue-focused web crawler identifies and downloads any document (Web pages, word, pdf or PowerPoint documents) about the issue found on the Internet; typically around 10-20 thousand.

The documents are then analysed for references. So if a document, created by organisation X, refers organisation Y in the context we are focusing on, then we take it that organisation X deems organisation Y relevant to the issue. It also on average means that organisation Y, to some extent, influence organisation X on the particular issue.

After some consolidation and statistical filtering we end up with set of interlinked stakeholders; typically 1000 +/- 500. These stakeholders constitute a body of stakeholders whose relevance to the issue can be substantiated.

Using well known mathematical procedures we then calculate metrics of interest; mostly influence metrics.

Table 1 - Issue Influence Index™ (above)

Issue Influence Index™ is a generic measure of influence. It measures both direct and indirect influence and is calculated like a citation index.

The scale is linear, ranging from 1 (one) and upwards. An index of 1 can be interpreted as “no particular influence”. A stakeholder with an index of 4 can be interpreted as having twice the influence as someone who has an index of 2.

An organisation with no particular influence on the issue in focus has an Issue Influence Index™ value of 1.

Figure 1 (above) shows how organisations from Table 1 reference each other. The direction of the arrow shows the reference. The influence is consequently the other way.

The size of the dot representing each organisation is proportional to their total influence.
Three of the organisations, Barclays, Heart of the City and Investis; are not referenced by the rest of the group (nor vice versa). These 3 organisations thus derive their influence from organisations with less influence than those listed in Table 1.

Table 2 - Information Influence (Above)

Table 2 shows a metric popular in network analysis (where it is often called “betweenness centrality”). We refer to it as “Information Influence” because it gives a good indication of how central a stakeholder is to the distribution of information about the issue in focus.

News media and others with an editorial role (collection of information and then redistributing it in edited form to a large audience) usually have a high Information Influence.

Information Influence is not as good a measure of “real” influence as the Issue Influence Index™ in Table 1, but it shows who is a supplier of information about the issue to a large audience (directly or indirectly).

Figure 2 (above) shows how the organisations in Table 2 reference each other.

Unlike Figure 1 we here have a situation where all the top organisations reference each other.

Again the size of the dot representing each organisation is proportional to their real influence as listed in Table 1.

Contact us if you are interested in the full analysis.

Issue Influence Index™ is a trademark of Onalytica Ltd.

Who (really) influence the debate on Wal-Mart?

In a recent article, The New York Times described how Walmart, a giant retailer, is under fire from pressure groups because of “its wages, health insurance and treatment of workers”.

According to the article, Walmart has hired leading political PR experts and Edelman, a PR firm, to help them in the battle for hearts and minds.

The article also singles out two pressure groups, “Wal-Mart Watch” and “Wake Up Wal-Mart”, as being one notch more professional than the rest.

I decided to try and quantify how much influence the two groups have on the dicussion of Walmart.

To find out who influence the debate on Walmart, I downloaded what I found to be free and publicly available on the Internet that made references to Walmart. I employed some simple restrictions but ended up with just under 20.000 documents, articles and web pages. (Hereafter commonly referred to as documents).

The documents where then analysed for references. A reference is, for example, when The New York Times makes a reference to Financial Times or Walmart Watch. The reference can be a textual reference or a link.

In the documents 35,464 entities (companies, organisations, government agencies, individuals, etc.) made 494,421 such references to each other in a context relating to Walmart. After disregarding self-references and a bit of noise filtering I was left with a total 1781 entities that had some direct or indirect influence. (Hereafter referred to as stakeholders)

The results of the calculations are summarised in the Table 1 and 2.

Table 1 shows those stakeholders who were referenced by most other stakeholders. From the table we can see that The New York Times was referenced by 87 of the other 1780 stakeholders in the documents analysed.

One of the pressure groups from the New York Times article, Walmart Watch, is among the top 25 shown in Table 1. It’s number 16 (tied with Salon) and referenced by 36 of the other stakeholders. The other pressure group, Wake Up Walmart, is number 30 (Tied with Fast Company and others; 25 references) and thus just outside Table 1.

So from this measure it seems that the pressure groups have some influence, but as we shall see, things are not what they seem.

The numbers shown in Table 1 are often used as a measure of authority, but in fact the quality of the measure is not great.

There are two main reasons why this measure is not great. One is that it assumes that the sources referring are of equal weight: a mention in my dog’s blog count for as much as a mention in NY Times. Except for my dog, no one agrees with that.

The other problem is that the metric fails to measure indirect influence.

The reason why I have included it is that is a metric still use by some search engines (e.g. Technorati) as a measure of authority.

Precise measurement of influence is somewhat trickier.

In the academic world it has long been a tradition to calculate the influence of academic journals and universities based on something called citation-indexes. (See what Wikipedia writes about citation indexes).

This kind of influence measurement is based on the conjecture that if an author of an academic paper cites another academic paper as relevant to the context, then the citing author is, to some degree, influenced by the article being cited (and its author).

To put it more formally, if person A cites person B, then A’s contribution to B’s influence can be expressed as the total influence of A divided by the number of people A cite in the context.

So if A’s influence is 2 and A cite B and 3 others (4 in total) in the relevant context, then A’s contribution to B’s influence is ½ (2 divided by 4).

A’s total influence is the sum of all the contributions made in this way.

So, the more people you cite as relevant to a context, the less influence each of them has over you.

Table 2 shows the Issue Influence Index™. The metric can be directly interpreted as influence (or structural influence).

It is calculated in the same manner as the academic citation indexes described earlier and is a very precise measure of influence because it does not treat every one as equally influential and it takes indirect influence into account.

The Issue Influence Index™ is linear (ranging from 1 and upwards) so a metric of 4 can be interpreted as “twice the influence” as 2.

Table 2 shows that Walmart is the most influential when it comes to the topic if Walmart with New York Times and Yahoo as close seconds.

As number 3 and 4 on the list we find Walmart Watch and Wake Up Walmart; the two pressure groups whose influence we set out to determine. We can see that they have about the same influence (each of them) as Reuters.

In conclusion we can see that the two pressure groups whose influence we set out to investigate have substantially more influence than their number of citations would initially lead to believe.

For more information on measuring influence, see our White Paper.

Measuring influence

We have just released a White Paper titled

"Measuring the influence of stakeholders on an issue, cause, brand, institution, or company."

It's about on how a new type of stakeholder analysis is helping marketers, policy makers and public relations experts win the battle for awareness and influence.

Download it here to read it.