I thoroughly enjoyed the event – except that it overran and I had to miss the Q&A panel to run and get a flight. That aside, it’s really gratifying to watch a large-ish vendor execute on marketing strategy with a single vision. I’ve long admired Informatica’s marketing, having worked alongside a couple of their ex-senior marketing folks and seen the disciplines they learned at Informatica. And it’s clear that the company is focussed, knows who it sells to and why they buy.

But I thought I’d share a couple of observations that reminded me that the Informatica marketing machine is not infallible.

On Audience Strata Mismatch

Audience strata mismatch is the term I use to denote when a vendor is saying all the right things to the all the wrong people. Much of the content from the event centred around the business pains created by a poor data integration strategy and, obviously, the benefits that would flow to the business if the data integration strategy were fixed with, say, Informatica’s products. All good.

As marketeers, we’ve all seen this before. And we’re used to sitting through slide after slide of business benefit bingo which is usually pretty boring. Except that Informatica is able to point to case studies with quantifiable benefit and ROI which makes it slightly more interesting. Again, all good. The problem came during the excellent presentation by John Radcliffe of Gartner where he asked the audience to define their role in their company as either on the business or the technical side. I was surprised to see that, in John Radcliffe’s words, “That’s about 99% of you on the technical side.”. Hmm I thought. So Informatica built an excellent agenda, excellent speakers and excellent content, but perhaps not targeted completely accurately. Don’t get me wrong, there was technical content during the day and I don’t want to suggest that all people ‘on the technical side’ do not care about the benefit of the technology to their business. But, I’m sure they were expecting a different balance.

On Evangelism

Long, long ago when the Internet was just starting to be considered as a platform across which business to business commerce could be transacted, I went on a European speaking tour promoting a reliable computer to computer messaging product. I remember getting on my feet in front of audiences from London to Prague and from Paris to Milan evangelising how this messaging product would revolutionise the way business was conducted. I told them that the very fabric of the way business to business trade was conducted would change, I told them that they should show vision and leadership in their market and use our product to gain the march on their competition. I used examples of similar paradigm shifts from the past that companies had used to disrupt their markets. I implored them for the sake of the families to grab the wave and not miss out on this opportunity. Well perhaps I didn’t go quite that far, but I certainly played up the role that B2B e-commerce would have.

Was I wrong? No. Was I premature? Well, only by about 5 years.

 On BIG-DATA

Why am I sharing this fact from my past? Well BIG-DATA appears to be Informatica’s ‘reliable computer to computer messaging’ and social networks, Informatica’s ‘B2B commerce’. Informatica sees the exploitation of social media and the resultant mining of the massive amounts of data by-product of our twittering, liking and plus oneing; as a massive potential opportunity for their clients and prospects. You see, for a company that is all about data integration and that has a solution for every conceivable current (budgeted) data integration project type, where do you go next? The answer appears to be that you get into BIG-DATA integration, ahead of your competition, but also apparently ahead of demand.

So why do I say ‘ahead of demand’? Well because the level of evangelistic fervour with which Informatica approached this massive, future, potential opportunity, reminded me of my good self all those years ago. When the (excellent by the way) presenter drew a comparison between the competitive motivation the United States used to start its space program in the late ’50s and the current competitive market opportunity that presented itself to the audience, I sensed we might be ahead of the curve.

That’s not to say that there isn’t an opportunity right now for organisations looking to mine the BIG-DATA created by social media mechanisms to interact with the customers. I’m sure there is. But for the majority of folks in the audience this is likely to be some way off and, as a result, some way down their list of priorities. It’s clear that BIG-DATA is going to give data integration vendors headaches and opportunities in equal measure, but it was also clear that the demand for BIG-DATA solutions is nascent at best.

Anyway, an excellent and educational event and I’ll be back in 5 years to check on just how ahead of the curve Informatica is/was.

Danny Goodall

Disclosure: I advise an organisation in the MDM space and Informatica acquired a company that I previously advised.

Related posts that you might also be interested in...

REPAMATron automates my REPAMA competitive marketing intelligence methodology and is currently in Alpha. I’m using the ESB market, a market I know well, to help to tune the algorithms at the heart of REPAMATron.

In my previous entry I looked at 3 Enterprise Service Bus vendors’ marketing strategies (WSO2, Talend and FuseSource) and compared them to the computed market mean. I’ve now added another 6 ESB vendors to the study from the long list of ESB vendors produced by Gartner. These new vendors are TmaxSoft, JBOSS, Axway, Seeburger, JBOSS, InterSystems and iWay. So the complete list of vendors and products currently under scrutiny is:

I’ve also categorised the list into open source and closed source ESBs. This has allowed me to pull together Marketing Element Distribution (MED) charts that show the difference in marketing strategy between the average open source strategy and the average closed source strategy.

Which features are important for Open Source ESBs versus Closed Source ESBs versus the ESB average?

The PIPESCOM Marketing Element Distribution (MED) chart below shows the features that are promoted most prominently by vendors in the ESB space. Each of the 3 chart series shows a different market mean. The ESB market mean shows the 12 vendor average of all of the vendors in the table above. The closed source ESB vendor average features the average of the 7 closed source vendor results whilst the open source ESB vendor series aggregates the results of 5 open source vendors.

This is an interesting set of results for a couple of reasons. Firstly because there is a great deal of similarity between the open source and closed source vendors. Taking an ESB to market, be it closed source or open source, appears to be about focusing on communicating the same basic set of features. Interfaces, process and management features are all promoted strongly in keeping with the very nature of an ESB.

However, a number of differences (differentiation) exist. As one might expect the open source ESB vendors place extra emphasis on the packaging of their product as well as the commercial characteristics. There is also an additional focus on ease-of-use by closed source vendors although the reason for this is not immediately clear. It appears that closed source vendors believe that stressing the ease-of-use characteristics of their products is more important than their open source contemporaries.

Which value propositions (benefits) are important for Open Source ESBs versus Closed Source ESBs versus the ESB average?

The MITICOR Marketing Element Distribution (MED) chart below shows the value proposition (benefit) categories that the vendors in the study stress in their outbound communication. The chart series are as described in the PIPESCOM chart above.

Again, this is an interesting set of results. It appears that when it comes to expressing the value (or benefit) that the vendors feel they deliver to their customers there is significantly more differentiation between open and closed source ESBs than with the PIPESCOM features chart above.

Also of interest is the Cost value proposition. As you’d expect this is where a vendor feels they can save, reduce, offset, mitigate of remove cost for their customers. One would naturally think that an open source ESB proposition might focus heavily on the cost benefits of an open source solution over a comparable closed source ESB. The chart above bears this out in that the open source Cost proposition scores higher than the closed source equivalent. However the difference is very small and the Cost value proposition for closed source vendors already scores strongly. This would suggest that with their closed source ESB counterparts already focussing on the cost benefits of their solution, it will be difficult for the open source vendors to differentiate and sell using cost alone as a proposition.

In my next blog I’ll share the results of a direct comparison between an open source ESB vendor and a closed source competitor. I’ll also continue to add vendors and additional tests to these studies and will share the results in future blogs.

Please post any feedback you may have in the comment section below or contact me here.

Danny Goodall

Related posts that you might also be interested in...

The previous post described what REPAMA and REPAMATron are all about so if you’ve landed here and have no idea what this is, then I’d recommend that you read the first post in this series. Having taken a look at how a section of the Enterprise Service Bus market (WSO2, FuseSource and Talend) communicates benefits or value propositions, I’m now going to share the research into the product features that each vendor communicates most prominently. To allow a comparison to be made between the vendors and to the market mean (the average strategy of all vendors in the segment), I first have to examine how each vendor communicates product features and determine which of the PIPESCOM categories the feature fits into.

What is PIPESCOM?

The PIPESCOM classification allows one vendor’s marketing copy to be compared to another’s with respect to how they communicate product features. This is achieved by categorising the features that are described in the vendor’s marketing copy into the following broad categories. Packaging, Interfaces, Process, Ease of use, Speed, Commercial, Operational, Management. So for example, if a vendor describes their product as having “the ability to interface with other systems” then that would suggest they are making a claim about the product’s Interfaces. If a vendor talked about their product being available as a service then this would relate to the Packaging category of PIPESCOM. Similarly if the product is described as ‘inexpensive’ or ‘commercial-off-the-shelf software’ then that would fall into the Commercial category. You can read more about PIPESCOM here.

PIPESCOM (Product features)

The following series of charts shows the relative support for each of the PIPESCOM product feature categories found in the vendor’s marketing materials. The esb market mean displayed in the chart is the aggregate or average strategy of the 6 Enterprise Service Bus Vendors in the table below.

This market mean allows a comparison to be made between the results of WSO2, Talend and FuseSoruce and the ‘average’ approach vendors take to the Enterprise Service Bus market.

FuseSource

^*The chart below shows that FuseSource stresses the interfaces the product has to other technologies. This is not surprising since an ESB exists to interface to other technologies. That said, it is clear that FuseSource’s reliance on communicating Interface features is much higher than the esb market mean (the aggregate of the 6 ESB vendors above). FuseSource does not focus on ease of use, or at least not significantly enough to register with REPAMATron.

Talend

^*Talend’s product feature focus mirrors very closely the esb market mean. The exception is that Talend focuses significant effort on ease of use features – significantly more than the market mean.

WSO2

^*WSO2 shows an above average focus on packaging (the way the product is configured for sale/use) and the management features of the product. Similarly to FuseSource, WSO2 places little discernible focus on the ease-of-use features of its products.

FuseSource, Talend and WSO2

^*The combined chart shows how each vendor stresses different product features in their outbound marketing. The differentiation between them and to the esb marketing mean is shown by their relative commitment/lack of commitment to one of the PIPESCOM categories. It appears that differentiation exists between the vendors around how Packaging, Ease-of-use, Commercial and Management features are communicated.

In future posts I will expand on the vendors in this ESB study and will also share some additional MED and text analysis tests. I’d welcome feedback about vendors that should be included in these studies and indeed what market segments I should look at next. In addition to rounding out the ESB study I plan to look at high performance messaging and master data management as I am working on client projects in those areas at the moment. Please post feedback in the comment section below or contact me here. Danny Goodall ^* – This is alpha software. The results and conclusions are indicative at the moment and are in no way definitive.

Related posts that you might also be interested in...

If you’ve been following this blog you’ll know that I analyse the marketing tactics and strategies of high-tech vendors – mainly in the infrastructure software space. It’s…

…an emotional day for me as I share the first output from my automated competitive marketing intelligence gathering system – REPAMATron. Well perhaps it’s not that emotional, but it certainly represents a significant milestone in a project that has been in development in my ‘spare time’ for a little while now. Below you’ll find my first tentative steps at automated analysis of the Enterprise Service Bus market.

REPAMA is a research methodology that allows me to categorise and compare the marketing strategies of high-tech vendors inferred from the language they use to take their products to market. REPAMATron automates this manual methodology and generates competitive marketing intelligence that allows competitive differentiation to be visualised and benchmarked.

Before I share the first research results, here is some of the background alternatively you can go straight to the research.

It’s Alpha. It’s Alpha. It’s Alpha.

Taking inspiration from Google’s legendary ‘Beta’ logo that it uses so well to reflect a bit of geek chic onto their latest features, I feel the need to do the same. The difference is that one suspects that Google’s beta code is pretty much commercial quality with a beta label stuck on it for marketing purposes, but with me it’s important to say that this really is alpha software.

The technology and logic behind REPAMATron is at a very early stage. The rules that it uses to determine whether a vendor is following a certain strategy or not will be continually refined over time and as such it will take a little while until results stabilise enough to be ‘accurate’. That said the general inferences that it arrives at are broadly accurate today.

What’s the big idea?

This is competitive differentiation Visualised and Benchmarked.

Visualised - I wanted to create technology that would visualise the competitive differentiation between different vendors’ products within a given market segment. This is especially important in markets where little apparent differentiation exists between the competing vendors. Developing REPAMA taught me that even amongst vendors that have broadly similar products; the way they describe, promote and generally take their products to market can be very different indeed. I wanted to develop a system that could identify ‘hidden’ differentiation and present it graphically so that it could be easily understood.

Benchmarked – By grouping together the a number of vendors’ results it is possible to create an average set of results – I refer to this as the market mean. It is then possible to benchmark and compare each vendor’s strategy to this market mean as well as to specific competitors in the market segment. Understanding how your own strategy differs from the market mean or from a specific competitor can be a very valuable insight.

How does it work?

If I told you that I’d have to kill you. But in general terms, ScrewTinny, the technology behind REPAMATron collects examples of marketing communication from vendors – usually marketing and product collateral from web sites, press releases, emails, etc. This text is analysed using phrase matching, natural language processing and statistical likelihood methods to determine if a specific marketing strategy is present or if a specific claim is being made. All of these inferences are then processed as if being read by a human so text found on popular web pages ranks higher as does text near the top of a document as do multiple clusters of the same inference, etc.

There are currently two main types of analysis. MED Charts and TextAnalysis charts. TextAnalysis charts are not shown below but I will post more on them in a future blog entry.

What is an MED Chart?

The MED Chart (Market Element Distribution), is used to show the relative commitment a vendor has to a specific strategy. A score of 11^* is apportioned, or distributed across the various different marketing strategies in the chart. Vendors that focus their marketing communications on a small set of strategies are rewarded with a higher score. However vendors that water down their focus by aiming at many strategies will score less for each strategy.

Let’s take an example of a two vendors that are describing the value proposition (the benefits) that their respective products confer. Let’s say that vendor A makes one claim repeatedly that his product reduces the cost associated with some task or other. Vendor B also makes repeated claims that her product reduces cost, but in addition she also claims that her product will increase her customers’ revenue and will also reduce the risk of project failure.

Now both vendors make a value proposition to their audience associated with reducing cost but for vendor B her cost claims are watered down because she also makes additional claims about value propositions (or benefits).

What is MITICOR?

MITICOR is my acronym for defining the major categories of value proposition. Think of MITICOR as capturing the category of benefit that each vendor believes their product provides. Broadly speaking I have categorised business to business value propositions for infrastructure software into the following broad categories: Market, Income, Time, Institution, Cost, Operational, Risk. So a Market value proposition might be a claim that you will increase your customers’ market share or that they will get some form of competitive advantage whilst an Institutional value proposition might instead talk about delivering improved company image or reputation, increased shareholder value, better corporate governance, etc. You can read more about it here.

What PIPESCOM?

So if MITICOR captures the claimed benefits associated with a product then PIPESCOM captures the features that the vendor communicates most prominently. As I explained in this previous PIPESCOM blog entry, it’s difficult to compare vendors product marketing strategies without first trying to categorise the types of features that are promoted.

PIPESCOM allows one vendor’s marketing copy to be compared with another vendor’s by categorising the claimed features of the product into the following broad categories. Packaging, Interfaces, Process, Ease of use, Speed, Commercial, Operational, Management. So for example if a vendor describes their product as having “the ability to interface with other systems” then that would suggest they are making a claim about the product’s Interfaces. Similarly if the product is described as ‘inexpensive’ or ‘commercial-off-the-shelf software’ then that would fall into the Commercial category. You can read more about PIPESCOM here.

What’s Next?

I’ve only automated a couple of the 20+ different studies found in the manual REPAMA methodology so I will add those that can realistically be automated. In addition having access to statistical information about the marketing copy that each vendor uses opens up many more possibilities that can’t, or at least can’t realistically be done manually.

The Research

When looking at the vendor comparison below please bear in mind that I’m not looking at the actual capabilities of the products – this isn’t a product review. Instead I’m looking at the marketing claims each vendor makes for the capabilities of their products as embodied in their outbound marketing communications.

As I mentioned in a previous blog post I chose the ESB market to help hone and develop the algorithms behind REPAMATron because it is a market I know well. I will continue to add vendors from this long list of ESB vendors over time and ultimately will cover all market segments in infrastructure software, but for the moment the vendors and products included in the analysis are:

From that list, the Talend, FuseSource and WSO2 results are shown below. The esb market mean comparison that is shown on the charts is the notional average of all 6 ESB vendors from the table above.

The charts are interactive in that if you hover over or click the various datapoints you should be able to see the specific values.

MITICOR – Value Propositions (Benefits)

The following series of charts show the value propositions or benefits found in each vendor’s marketing copy. The MITICOR charts show the relative importance that each vendor attaches to that category.

FuseSource

^**In the chart below it is clear that FuseSource stresses the cost and operational benefits of its software. No other value proposition was found that scored highly enough to be recorded in the chart. This shows a clear focus which interestingly, is quite a departure from the market mean of the 6 ESB vendors. Fuse Source’s focus on Cost is considerably higher than the market mean whilst their their claims about the operational benefits of the product is considerable below the market mean.

Talend

^**Talend’s value proposition appears to centre completely around Operational benefits. Any other value proposition that might appear in its marketing copy didn’t register highly enough to score on the chart. This type of result, where a only single value proposition is recorded is quite common amongst vendors that focus on features without linking them through to benefits that their customers might derive from the feature.

WSO2

^**Interestingly, WSO2′s value propostion most closely follows the market mean. The implied benefits for WSO2′s products are that they will save time, reduce cost and improve operational efficiencies.

FuseSource, Talend and WSO2

^**The combined value proposition chart for the three vendors shows the differentiation found in their marketing materials – differentiation between the 3 vendors in the chart as well as differentiation versus the the ESB market mean.

Well that’s the MITICOR value proposition charts, I’ll post the PIPESCOM product features analysis in a separate blog entry.

I’d welcome feedback about vendors that should be included in these studies and indeed what market segments I should look at next. In addition to rounding out the ESB study I plan to look at high performance messaging and master data management as I am working on client projects in those areas at the moment.

Please post feedback in the comment section below or contact me here.

Danny Goodall

^* – Why 11? Everyone knows that the best amplifiers go up to 11.
^** – This is alpha software. The results and conclusions are indicative at the moment and are in no way definitive.

Related posts that you might also be interested in...

In a previous post I mentioned that I would follow-up to with a calculation for how much jargon WSO2 uses in its ESB marketing when compared to Talend and FuseSource. I’ve now calculated the Arcanicity Index score for WSO2′s ESB marketing copy; and whilst not quite being the answer to life, the universe and everything, it’s not far off – see below.

The Arcanicity Index looks to measure how much jargon a text contains and by extension how much specialist knowledge a reader would need to posses in order to understand it. I’d previously scored both Talend and FuseSource which had shown that Talend’s marketing copy was much less arcane than FuseSource’s but oddly did require a much higher level of grade education to understand it.

It should be said that in a sector of the market where all three of these vendors are presumably targeting overtly technical people, having highly arcane content is not necessarily a bad thing. In fact it’s almost self-selecting. If you’re audience can’t understand your marketing copy then they’re probably not you’re audience. But we do have to recognise that we might be excluding some from understanding the text we write.

Anyway, the results are in and first I have to say that I’ve had to adjust the previous score for FuseSource. I’m now using a different algorithm to detect the number of sentences in a text as well as new code to more accurately estimate the number of syllables in a word. This new algorithm counts less paragraphs for FuseSource which in turn increases the Arcanicity ratings ( as it is a function of acronym density per paragraph ).

And as the table below shows, WSO2 scores somewhere in the middle with an arcanicity score of 45.07 and an arcanicity rating of ‘moderately-high‘ for its ESB marketing copy. The sample text came from WSO2′s ESB product PDF as opposed to their web site because there wasn’t really enough of a sample there.

A couple of caveats for these results. Firstly, as mentioned above WSO2′s text came from a product PDF whereas for both Talend and FuseSource I used the ESB product page from their respective web sites. I’d expect product collateral like a mult-page PDF to be more technically-focused than a general product web page and so in all likelihood WSO2′s generic marketing materials would produce a lower result . Secondly, the WSO2 name is being picked up as an acronym in these results which again is certainly increasing the perceived acronym density of WSO2 here.

I’ve actually pushed some code up to Google’s cloud-hosted appspot.com domain that should allow others to calculate the Arcanicity of their own text – as well as a host of other text statistics. If you’re interested it’s at http://arcanicity.appspot.com. But be gentle as it’s alpha software and I’m not a real programmer!

Danny Goodall

P.S. Here are the respective acronym frequency distribution charts. Click to enlarge.

WSO2

FuseSource

Talend

Related posts that you might also be interested in...

I’m using this study to road-test Project REPAMATron – an automated system that infers a vendors’s marketing strategy from their marketing materials. I’ve decided to add WSO2 into the mix for a number of reasons. The main reason being that I spoke to a representative of one of the other ESB vendors on my long list who asked me “Are you going to be looking at WSO2?”. Which was good enough for me. If that vendor was concerned enough about WSO2 to ask if they’d be on the list, well, then they have to go on the list.

Secondly, in addition to Talend and FuseSource, it will be a third vendor that is building Open Source technology that relies on the Apache stack in some way. It will be good to see how ‘similar‘ products are able to differentiate themselves through their marketing.

Lastly, I really like the marketing approach that the company appears to be taking. It has gone for a “Paradigm-Shift – it doesn’t have to be this way” strategy. Ironically, when we introduced the ESB category at Sonic Software, that was the same approach we went for then – under the banner of

“Changing the Economics of Integration”.

Also, in step with the Sonic Strategy, WSO2 is trotting out the…

“We started from scratch therefore we don’t have the same bloated, overweight stacks as our older (and massively more succesful BTW) competitors”.

…line too. Sonic, and other start-up ESB vendors also levelled that differentiator at the established integration vendors of the day.

But underneath that the company does really seem to have embraced the energy, vibrancy and indeed the new paradigm of the Open Source world. Proof of their willingness to challenge convention is their CEO Sanjiva Weerawarana pictured on their web site standing on his head. Which at least challenges the age-old paradigm of standing on your feet.

So hopefully we’ll see how much the paradigm has shifted. The people there certainly seem to have good provenance. Andwhilst this process is not really about the ESB vendors but more about helping me test a piece of software to see if it is inferring the right meaning behind the marketing language that the vendors use to take their products to market, I can’t help thinking that I’d really like to see how WSO2′s fresh approach differentiates it in the market.

Just as I did with Talend and Fuse Source, I’ll run WSO2′s marketing copy through the Arcanicity Index later to see how much jargon they use in their marketing copy.

I’ll keep you posted.

Danny Goodall

P.S. If there any other vendors you think should be added to the project – let me know.

Related posts that you might also be interested in...

EDIT: Here’s a link to the Arcanicity Index Calculator I mention lower down in this post.

I stumbled across this site the other day when looking for information on the frequency of words used in common English. The site is the brainchild of Jonathan Harris and is wonderfully minimalistic. It almost feels as though this simple Flash plugin were an installation in a museum’s white room. A real lesson to us “more is more” types that indeed, less is more.

Anyway, take a look at wordcount.org and marvel at where your favourite words, your first name, your last name or your company name all fit into the rank of words. But be warned, it’s sister site Querycount watches you and ranks the frequency of searches in wordcount. Go on, have a guess how many expletives make it into the top ten of wordcount searches. Well the answer is three. And they’re three of the worst, or best, depending on your point of view.

I’ve been thinking about readability and ‘Arcanicity’ recently and have been trying to rate how easy a piece of text is to comprehend. Wordcount has made me think that a good place to start might be to look at how frequently each of the words in the text appears in common English usage. After all, if a sentence were full of very uncommon words it would surely mean that a reader would need greater knowledge / experience to comprehend it.

Hmm. Something to ponder.

Danny Goodall

Related posts that you might also be interested in...

I promised that I would share some real-world examples of how high-tech vendors’ marketing copy scores in my Arcanicity Index. So first up is some text from the respective web sites of Talend and FuseSource – two ESB vendors that I’m looking at in detail to test project REPAMATron.

As a reminder the Arcanicity Index looks to rate how much technical, or domain knowledge a reader would need to posses in order to understand a given piece of text. At the moment it works from the density of the abbreviations and acronyms found in the text. Future version may also look for technical concepts as well. The equation and MUCH more information on the Arcanicity Index can be found here or a slimmed down version is here.

The headline figures are as follows:

The average Grade Index here is taken from the average of Flesh-Kincaid, Gunning-Fog, Coleman-Liau, Smog and the Automated Reading Index all of which aim to calculate the number of years of study someone would need in order to be able to comprehend the text from a structural point of view. What this tells us is that on average the structure of the text on the FuseSource site is much (nearly 5 school grades) easier to read.

However, if we look at my Arcanicity Index score (which is still a working draft and not finalised yet!) for the respective texts it shows that the FuseSource copy features a far higher degree of arcane concepts when compared to the Talend copy. The Arcanicity Index has a notional maximum of 100 although it is possible to score higher than 100. FuseSource scores 72.76 on that scale which makes the concepts contained within very complex.

What all this means is that the FuseSource web copy is likely to exclude non-expert readers and suggests that FuseSource is aiming at a VERY technically savvy audience. Talend’s web copy, whilst still containing very complex concepts is written in a narrative way with long hand explanations of concepts instead of using abbreviations and acronyms. Who has it right? Well I guess time and their relative success will tell.

I guess this goes to show that there is a significant difference between grade reading age and the degree of arcane, technical concepts within a text. This is why I decided to introduce the Arcanicity Index in the first place so I’m pleased it’s adding an extra layer of insight.

This just leaves me to announce that the current leader in the Arcanicity Index “Geek of Geeks” award for the most acronym-dense marketing copy is FuseSource with a score of 72.76.

Danny Goodall

For those masochists, insomniacs or natural language processing lovers out there; the statistics relating to the two vendors’ texts are shown below – in GREAT detail.

Text Structure Statistics

Here we see the statistics for the sentence and word structure. It’s clear from this table below that Talend uses longer and more complicated words when describing it’s ESB product offer.

Traditional Readability Index Scores

The table below shows the scores for Flesch-Kincaid, Gunning-Fog, Coleman-Liau, SMOG and the Automatic Reading Index. As mentioned above these indexes each attempts to predict the number of grade years of education that would be required to comprehend the text. More information can be found here.

Acronym Counts

The table below shows that statistics relating to acronyms. There are two sets of counts for each statistic – one for unique acronyms within the text and one that counts duplicates.

Syllable counts

The table below shows the syllable counts for the words within each of the texts. The first line that begins [(1,93), (2,74)...] shows that there were 93 x 1 syllable words and 74 x 2 syllable words, etc. in Talend’s web copy. Again it is clear that despite the amount of words in the FuseSource sample being nearly double that in the Talend sample, Talend has very nearly as many 3 syllable words as FuseSource and only one less 4 sylaable word. Talend uses longer, more complex words to describe its offer.

Talend ESB Web Copy

Here is a sample of text taken from the ESB product page from Talend’s web site.

Talend ESB is a lightweight, modular solution for various enterprise integration requirements. An ESB provides a combination of service enablement, messaging transport, routing, and mediation to address a wide variety of enterprise integration challenges.  Talend ESB supports a broad set of standard enterprise integration patterns (EIPs) making it an excellent solution regardless of topology or architecture.  Moreover, Talend ESB is open and standards-based and therefore is widely interoperable with existing and legacy infrastructure components.

Talend Service Factory is designed for developers that are looking for rich service enablement capability without the additional overhead of a full-fledged ESB.  Talend Service Factory supports both Web services and REST based service development giving developers the freedom to choose the method most suitable for their existing environment.

Talend Integration Factory Community Edition simplifies integration tasks and makes users more productive in their integration projects. Talend Integration Factory is based on the open source Apache Camel project and is built on a lightweight, component-based architecture that supports a variety of integration protocols, data formats and programming languages.  It comes “out-of-the-box” with pre-configured runtime packages ready for deployment in environments including OSGi, Apache Tomcat, JEE servers or standalone JVM.

Learn about Talend’s application integration products:
Talend ESB, a flexible solution for complex distributed integration projects.
Talend Service Factory, an open, lightweight and flexible architecture  for service enablement.
Talend Integration Factory, a fully supported, production-ready mediation and routing product based on Apache Camel that uses Enterprise Integration Patterns to improve developer productivity

FuseSource Web Copy

Here is a similar sample taken from the product page for the FuseSource Fuse ESB.

The Open Source ESB for the Enterprise.Fuse ESB is certified, productized and fully supported by the people who wrote the code. Fuse ESB has a pluggable architecture that allows organizations to use their preferred service solutions in their SOA. Any standard JBI or OSGi-compliant service engine or binding component – including BPEL, XSLT or JMX engines – may be deployed to a Fuse ESB container, and Fuse ESB components may be deployed to other ESBs.

Key Features and Benefits
Feature

Standards-based design – JBI, OSGi, JMS, JCA, JMX, BPEL support Apache license – Fuse ESB license is based on Apache’s open source license
Supports any app server – WebSphere, WebLogic, JBoss and many others
Productized – integrated, tested, and releases are versioned
Supported – training, consulting, and subscriptions are available
Backed by Apache leaders – Fuse employs key ServiceMix contributers

Benefit

Straightforward integration – reuse existing middleware components
Embeddable – can freely modify and redistribute the code
Flexible – use the development environment of your choice
Enterprise-class – can be used in large, mission-critical projects
Collaboration – comprehensive help for evaluation, development, deployment
Access to Apache – access to and influence over project roadmaps
Interested in more detail? Read the data sheet.

Ready to Get Started?

Download

Download Fuse ESB (based on Apache ServiceMix)Looking for Fuse ESB 3? Downloads of earlier versions are available here.
Unix/Mac (81 MB, TAR.GZ)
Windows (82 MB, ZIP)
Source (4.5 MB, ZIP)
Waiting for your download to complete? Why not check out how FuseSource subscription, training and consulting for Apache ServiceMix can help you be as successful as possible making FuseSource a part of your enterprise?

Downloading the product is just the beginning!As a community member, you can:
ask and answer Apache ServiceMix and JBI questions in the community forums
document your investigation of the Apache ServiceMix product in the wiki
use the community repository browser to access builds in Subversion
receive discounts for Apache conferencesand much more…Join our community today!

Demo Videos
To run the demos yourself, first review the slide presentation then download the code:
These videos are .mov files which require QuickTime to view.
Building, installing and running Fuse ESB 4 (screencast): In this session, Ade shows how to build, install and run Fuse ESB 4. He also shows how to “prune” the installation so that FUSE ESB only loads up the modules you need for your work.
Deploying a POJO into Fuse ESB 4 (screencast): In this session, Ade gives an overview of his development environment, and shows how to deploy and share your Java POJOs into Fuse ESB 4. He shows how to integrate your logging with the Fuse ESB 4 logging system, and, how to read your own custom configuration data using properties files and the OSGi configuration service.
Deploying a Web Service into Fuse ESB 4 (screencast): In this session, Ade shows how to create a SOAP/HTTP web service using JAX-WS and deploy it into the Fuse ESB 4 runtime. He also shows how to tie this new component in with the shared POJO developed previously.
Deploying an Enterprise Integration Pattern into Fuse ESB 4 (screencast): In this session, Ade shows how to create and deploy enterprise integration patterns. Using the Apache Camel DSL, he creates a file- and FTP-poller to read in plain-text CSV files, saves them to a persistent queue using JMS, and then splits the batch data line-by-line for individual processing.
Deploying a JBI component into Fuse ESB 4 (screencast): In this session, Ade shows how to take JBI solutions developed and deployed in Fuse ESB 3 and deploy them in the Fuse ESB 4 container.

And that, was all she wrote.

Related posts that you might also be interested in...

I’ve been posting recently about my desire to rate the readability of high-tech vendors’ marketing copy. There are many existing readability indexes. Each of which looks to calculate the ease at which the structure of the sentences/words can be comprehended. But none that I’ve found tries to gauge how technically difficult it is to understand the concepts contained within the text.

I’m interested in this because as part of project REPAMATron, I want to be able to compare various vendors’ marketing copy. It also gives me an insight in to what audience strata (IT Technical, IT Business or Business) a vendor is targetting. Anyway, I’ve settled on a calculation that seems to hit all of the design criteria I set for a readability index that takes into account the level of domain knowledge required to interpret text.

And here it is…

Let me explain how I got to this first draft and exactly what that calculation means

But first let me recap on the issue I am looking to address. If, during the course of my analysis I were to find the following text on a vendor’s web site, product PDF or PowerPoint presentation…

ACME’s IJK-compliant WooshBang product provides developers with an integrated IDE for developing RTY, JKL and ERT applications. ASDF support is provided through a plugin which also serves as a POL generator. Developers that previously used NMB or XCV will find the environment totally familiar to them. Security is enhanced through the addition of RTY scripts which when used in conjunction with the POL generator ensures unauthorised access is not allowed. Finally, upgrading from previous versions of WhooshBang is simple. Just install the new version and old UIO repositories will be upgraded automatically.

…I might come to the conclusion that the text is pretty much impenetrable. Because whilst it’s relatively simple to interpret the physical structure of the text, unless the reader understands those acronyms, the meaning of the text remains hidden. And before you rush out to buy a WhooshBang, I’m sorry to disappoint you but it’s fictional.

But just how difficult is it to understand that text? Is it aimed solely at an audience with the domain knowledge to be able to understand it? Does the vendor, ACME in this case, realise that they are excluding readers from understanding the content of their web site? And is the text any more or less difficult to understand than text used by ACME’s major competitors? Well these are the questions I am looking to my Arcanicity Index to answer.

How might we calculate the degree of arcane content in a text like this? Well, let’s look at some general text statistics from the text above and see if we can identify some metrics we could use.

General Text Statistics

These statistics give a good indication as to the complexity of the structure of the text itself. These are the metrics  used by traditional reading indexes such as Flesch-Kincaid, Gunning-Fog, Coleman-Liau, etc.

General Readability Index Statistics

And here are some of the general text readability indexes that look to estimate the amount of education someone would need in order to be able to comprehend that text. These indexes concentrate only on the physical structure of the paragraph/sentences/syllables and words.

Acronym Density Statistics

So what if we were to look at the acronyms and abbreviations contained within the text. What if we tried to compute the density of these arcane words? Would that help us to determine whether the content was aimed at someone with specific technical knowledge? Well here are the key acronym statistics for the text above.

Which Measurements do we need to calculate the Index?

In designing an algorithm that could gauge how complex a text is to understand from a domain knowledge perspective, I’ve concluded that I need three key metrics in my calculation.

1. NAS - NON-UNIQUE ACRONYMS PER SENTENCE
   • The density of the total number of acronyms per sentence (or the total acronyms in the text divided by the number of sentences in the text)
2. UAS - UNIQUE ACRONYMS PER SENTENCE
   • The density of the unique acronyms per sentence (or the count of unique acronyms in the text divided by the number of sentences in the text)
3. PAS - PEAK UNIQUE ACRONYMS IN A SINGLE SENTENCE
   • The peak number of unique acronyms found in a single sentence

The combination of these three metrics gives me a good indication as to the number, average density and peak density of the acronyms within a specific text. The question then is how to accommodate these in some form of index. To allow the index to be easily interpreted I want it to max out at 100. So a score of 100 would mean that total domain specific knowledge would be required to comprehend it. However if I reserve the 100 score for a text that is completely full of acronyms (i.e. all 100 words of a 100 word text are different acronyms) then scores for less acronym-dense text would be far too small to measure. The scores would be disproportionately skewed towards zero.

I needed a way to allow me to reserve a score of 100 for text that is aimed at experts only, but also allow for less acronym-dense text to receive a representative score too. I therefore needed to find a density point at which the text is already in effect “complex enough” and this point would become 100. I’m looking for a point where as more acronyms are added, the sentence doesn’t really get more complex to understand as it already requires a qualified reader. This would mean that it would be possible to score higher than 100 – much higher in fact, but 100 would be the notional maximum. Any score higher than 100 means that the density of acronyms is greater but the meaning is equally obscure.

So what would a score of 100 mean?

I felt that there must be a limit for acronyms per sentence density (#1 above) across a text extract where increasing that density doesn’t really exclude any non-qualified readers as the text is already at expert level. Likewise there must be a density of unique acronyms per sentence (#2 above) where readers are not excluded. Lastly, the measure of the peak number of unique acronyms in any one sentence (#3 above) would give a good indication of a text that was meant for expert eyes.

So I then did a bit of web crawling and looked at the typical count and density of acronyms on a number of infrastructure software vendors’ web sites (my area of focus and expertise). I rated the ‘arcanicity’ (the degree of arcane content) of those sites arbitrarily myself. I read the marketing copy. I counted the sentences and acronyms, worked out the densities and counts for 1,2 and 3 above and looking at my results across a number of sites, I ended up with the following measurements.

• 2.8 x NON-UNIQUE ACRONYMS PER SENTENCE
• 1.7 x UNIQUE ACRONYMS PER SENTENCE
• 4.0 x PEAK UNIQUE ACRONYMS IN SINGLE SENTENCE

Why 2.8, 1.7 and 4.0?

Well those numbers were the point at which I felt the text was already obscure enough. Or put another way, I felt that a sentence is for ‘expert eyes only’ if it contains an average of 2.8 acronyms per sentence (when looking at the total count of acronyms – not just unique acronyms) AND has 1.7 unique acronyms per sentence AND features one sentence that has 4 unique acronyms within it. This therefore is maximum arcanicity. I’ve capped the peak unique acronyms at 4 because again I felt that if a single sentence contained 4 unique acronyms, it was already expecting a degree of expertise from the reader in order for them to be able to understand it. Would a 5th acronym make it any less impenetrable to someone without domain knowledge? Probably not.

Creating the Acanicity Index Equation

Now I needed to put this into an equation that would get me a score of 100 if the text contained 2.8 non-unique acronyms per sentence, 1.7 unique acronyms per sentence and a peak of 4 unique acronyms in one sentence. So given that we want a maximum score to be 100, working back from there gives us something like.

100 = (1) x (2) x (3) x MAGIC NUMBER

or

100 = 3.8 x 2.7 x 4 x 2.4366

Note that I’ve added 1.0 to each of the density figures to ensure that they are always 1.0 or above. Otherwise density figures of less than 1 acronym per sentence would serve to lower the index and would create an artificial pivot point in the scoring around a density of 1.0. The 2.4366 is the magic number that when used with the the maximum vales of 1,2 and 3, creates a maximum score of 100. Or mathematically…

MAGIC NUMBER = 100 / (3.8 x 2.7 x 4)

And so this becomes…

Arcanicity = 2.4366 x NAS x UAS x PAS

As I mentioned above it’s possible to score greater than 100 but I’d contend that the sentence isn’t excluding any further readers at that point, as it is already arcane enough.

So how does ACME’s marketing copy above score?

Well the Goodall Arcanicity Index is 69.04, or…

(1 + 1.833) x (1 + 1.50) x 2.4366 x 4 = 69.04

Whether that is good or bad depends on who the text was aimed at. Without even seeing the text, a score of 69.04 tells me that the text relies on significant domain knowledge in order for the reader to understand it and I would hope that the vendor that wrote the text was aiming at the the IT Technical stratum.

Is that it?

No. The magic number and other ratios I’ve landed upon here in the first draft will need to be tested against more marketing copy to ensure that they reflect my own perceived level of obscurity. It will need to be fed very long passages and very short passages to see what min and max bounding I would need to put on the number of sentences in the input text. Currently I can see no value in processing less than 3 sentences – and 5 would probably be a more optimum minimum.

I expect that very long passages that are highly arcane is some sections but less so in others will produce an artificially low result. So I may need to adjust the equation to counter that in future drafts. I would also probably need a more rigorous, scientific approach to testing rather than my “finger in the air – yep it looks good to me” approach. Especially if this Arcanicity Index is to be used in the wild. But for now it’s a start and I’m quite pleased with the way it looks.

Over the next couple of weeks I’ll post some Arcanicity Index calculations for the ESB vendors that I’m using to tune project REPAMATron. And if you think my example text above for the fictional ACME corp was artificially complex, well I’ll show you some real-world examples that make that look like a Janet and John book.

Lastly, Why Arcanicity?

Well a number of reasons really.

• I like the word – it’s got a nice rhythm to it
• Google only returned 111 hits for it which means it isn’t a real word and it definitely won’t have been used before. I suspect the real term for the degree to which something is arcane or not is acranality, ancanity or probably arcaneness – but I’m not sure.
• I like the humour used in the names of other readability indexes such as Gunning-FOG and SMOG to illustrate that something was clouding the view of the meaning of the words. So I thought I’d inject a little, hopefully, obvious ironic humour myself. I mean, using a made up word with an obscure meaning and 5 syllables to represent an index that is designed to determine how arcane a piece of text is – well it just feels ‘right’.

Danny Goodall

Related posts that you might also be interested in...

[UPDATE - A more detailed description of the Goodall Arcanicity Index can be found here]

As I had mentioned before, I have been trying to create a mechanism to determine the amount of domain-specific knowledge a reader would need to posses in order to understand a text extract. The reason for this is that I am producing automated competitive-intelliegence studies using REPAMATron – a system I’m developing to infer marketing strategy from the marketing copy that high-tech vendors use to take their products to market. As part of those studies I want to be able to compare the technical content of different vendors’ marketing copy. Being able to estimate how much technical knowledge each vendor expects their audience to have in order to read their marketing copy, is important to me.

So I set out to look for an existing readability index that would provide a rating for how arcane a particular text is. I engaged with some experts in general readability indexes, but I couldn’t find anything. So I decided to produce something myself based on the density of acronyms and abbreviations in a given text.

And here it is…

It should be recognised that this is a first draft. The magic number of 2.4366 will probably need to be adjusted and I’ll also probably have to identify other arcane concepts in a text other than just acronyms and abbreviations. But this is a good start.

I’ll share how I arrived at this equation as well as the name in a future post. I will also provide examples using actual vendor’s marketing copy and perhaps award a prize for the most arcane marketing copy I can find.

But for time being, I give you the Goodall Arcanicity Index. You can read more about my motivation and the size of my ego here.

Danny Goodall

Related posts that you might also be interested in...