…Well strictly speaking that should be I’m creating a monster. You see, I’m in the process of automating our REPAMA methodology. Let me explain.
I’ve been using the REPAMA competitive research methodology for many years to understand vendors’ marketing strategies. Long before the formation of Lustratus in fact. It wasn’t a strict methodology back then. It was a set of relatively loose processes, measures and classifications that I used to ‘score’ various elements of a specific software vendor’s marketing strategy.
I used the process when I worked on the vendor side in various senior marketing positions. It allowed me to understand how I could differentiate my company’s technology from our competitors. And to differentiate, first I had to know how my competitors took their products to market.
I then used it to good effect in my marketing consulting engagements where I was able to advise software vendors how they should go about competing with other vendors in the market. I was able to provide an in-depth knowledge of their competitor’s go-to-market strategy – the sort of thing you only normally get by employing someone from a competitors’ organisation (after taking NDAs and employment contracts into account of course!) .
I’m sure that in an informal way most marketeers do the same. You’re reviewing a competitor’s marketing material and you’re trying to build a mental picture of the strategy behind the words they use. You ponder…
Which companies, in which segments do they target?
Which job roles are they looking to communicate with?
What product features do they promote most heavily?
What value do they claim to provide and how do they believe that they are different from their competition?
How similar are they to ‘us’?
How are they generating their sales leads?
Under the Lustratus brand I’ve been able to inject some structure and rigour into the process and I’ve given it a name – REPAMA or Reverse Engineered Positioning And Messaging Analysis. I now produce formal REPAMA Segment Analysis Studies for vendors who want to understand the marketing strategies and tactics of a vendor, a group of vendors or across a specific marketing segment. This they believe gives them a great advantage in their planning. All is good.
However, as the REPAMA Guide shows, it’s a complex process and it takes a quite some time to produce a formal REPAMA report. There is a great deal of planning, reading, analysis, categorisation, re-categorisation, scoring, ranking and documentation before the formal report can be produced. Add to this the fact that vendors’ strategies and tactics change all the time and it is quite a task for a single Marketing Analyst to keep abreast of a number of market segments.
So earlier this year I started to investigate how I might go about automating the process of producing this analysis. How, I pondered, might a piece of software be able to help me to ‘read’ and analyse examples of vendor marketing communication to produce the REPAMA reports? And so it was that Project REPAMATron (excuse the working title) was born.
Since then, I’ve been working through an external organisation to build the logic and process of my competitive analysis methodology into their software. The early results have been very encouraging and whilst it will be a little while before we’re going to be able to publish any research, we’re not far off being able to share some of the early results – hence this post.
The software is based around the techniques that search engines and other automated language based system use – namely NLP or Natural Language Processing. NLP strives to derive meaning from the otherwise apparently random collection of characters that computers use to represent text. NLP algorithms need to be ‘trained’ and the more language they see, the more accurately they can determine the meaning behind a collection of words or tokens. So at the moment there is a lot of ‘learning’ and tuning to be done.
Once the engine is designed, it is then a question of what form the published analysis might take. Currently I manually produce PowerPoint presentations and formal reports and I suspect that might be the first route I take with the automated analysis. But with the relative ease at which this information can be mined, it might be that I provide access to the analysis through some form of web-based on-demand search interface. It also means that I might be able to provide access to some of the base REPAMA information free of charge. And those are two characteristics of a search engine. And whilst I’m not talking about a general Google competitor here, it could be that a search engine for competitive marketing intelligence is getting a little closer.