Brand Monitoring and Text Mining Saved by the New Science of Consumer Perceptions and Interaction Outcomes

The brand intermediaries throw a curve ball while the early users of today’s immature text mining services are non-plussed.

Although I’m generally classified as a self-employed ‘business analyst’, my sub specialty is the critical review of technology-based product and services strategies. It’s been a problem getting the recruiter corps to wrap around my specialty, and I have to, by necessity, do my own marketing and outreach. Word of mouth, satisfied clients, and blog articles all help. I dwell on this because my latest madcap idea to re-focus brand monitoring as a decision support web hosted service for brand intermediaries toward the mid-market, and away from the brand owners, had stirred some ire in the industry.

“What do you know? What are your credentials? How did you come to this conclusion? What a conflagration.

Just because you put the word, ‘analyst’, after your name on a business card, does that make currency of your surmises and adjudications in themselves valuable? Of course not. Unless you carry the dubious distinction of working for the de facto pimps of the analyst industry – Gartner, Forrester, Giga, etc – you have to analyze and find the kernel of truth that is non-obvious, and oftentimes disruptive to the status quo and common beliefs  prevailing in the industry. Don’t get me wrong, I use the word pimps advisedly and gingerly; while these august agencies turn out a lot of paid for whore crap, you can’t beat their non-commissioned, self-sponsored reports and quantitative research that no lone-wolf analyst can match.

What do I know? Nothing special – I just try and call it as I see it. I have had a relatively undistinguished technology career with a few notable home runs. I write well, I express myself clearly, I can outline why I come to my conclusions and methods that led to those surmises and adjudications. I never graduated university, although I have one of the best interdisciplinary skill sets covering hardware and actual software systems and standards. This was all accidental, fate intervened.

What makes someone retain a bozo like me and say, ‘study this sector, look at what we are doing, or are proposing to do, and make a critical review’. Moi? Yes, moi. If a client has internal technical leadership calling a product play, I can be a cheap insurance policy to get some critical perspective on the pitfalls, or a clear headed referee for the go-ahead. Since I am not in the management or review chain, and my engagement is closed ended, I am no threat and my advice has often been disregarded. That’s my twinkle. I continue to get work because I more often than not catch a critical issue that was missed by those with an emotional or political attachment to a product or strategy or R&D path. It may be a mess when the client is in the midst of upheaval, but people go to other jobs and remember the good advice Wilensky  gave them back at old company X. The institutional investors and VC’s have been good to me, too.

Such a long introduction to this article on text mining of the public corpora for brand monitoring! There is a good reason, believe me. I have stirred up a lot of bad sentiment (ugh), by putting out the notion that the early entrants to the brand monitoring industry have delivered a weak set of services (currently, at least, Matt), and are likely over reporting their ostensible campaign volume, and have missed the fat part of the market that needs them the most, i.e., the mid-market intermediaries. They have missed the opportunity like all who have come before them in myriad industry’s because they followed rather than led in innovation, they saw brand owners as the low-hanging fruit because brand owners use roughly equivalent brand measurement services that are mirrored (not really) by the high-powered brand equity consulting industry.

As I have said repeatedly in my previous blog articles, and in my paid research for the previous EU telecommunications and internet client, there is no comparison between classical brand monitoring services and the new age early entrants of the text mining industry. Classical services by Arbitron and Gallup are far superior to text mining services for putting plans into action. Even Buzzmetrics, one of the true leaders of public corpus harvesting, is backed by ACN sector analysts for human perspective on what to do with collected data, how to conjoin with classical brand monitoring surveys, focus groups, etc.

Enough! Where, loud mouth Alan, did you ever get the idea for steering the industry towards multi-line distributors of the mid-market and retailer type brand intermediaries? Please! Well, I’ll tell you. And by the way, if any of the CMO’s and CEOs of the current crop of brand monitoring companies wants to keep going down the road they have chosen – no one is stopping them; there is no doubt that the cycle of CGM metrics services is still young, and that classical multi billion dollar agencies and new age textual harvesters will see a protracted period of consolidation.

I discovered the weaknesses of the current brand monitoring offerings by conducting interviews on behalf of my client. I interviewed the brand monitoring companies, of course, but I also interviewed their clients, their VC’s (off the record, and thanks to the Junior Partners that candidly expressed their bitter regret in ever getting involved with this benighted  sector), and by accident, I bumped into several of the brand owner’s down-line retailers and distributors – that’s about the time that the light bulb lit off above my head.

I love the part of the analyst business that involves thinking, discovering hidden relationships, opportunities, and synergies; the surveying part is boring, boring. Calls, calls, waiting, following up, etc., yuck. But this is how you put together data. As a one man show currently (I had an assistant back in the boom), it’s a bother. Sometimes, in the midst of the calls and filling out the survey forms, you get a tap on your sleepy head…huh! What did you say?

I was going through my calls, and somehow, one of the brand owners steered me towards one of their many retail distributors.  If I was paying attention, I might not have included the call in the list, but this one unintentional wrong turn was responsible for all that followed.

“What are your current options for monitoring brand and product reputation, are you using with, happy with, CGM services, blah….”. The answer that was to come, hit me with a iron rod:

“We are the victims of the branding decisions made by the manufacturers that we carry”. Full Stop.

I said, “Victims?” “Yes”, he continued, “Victims, recipients. We are always keeping a finger on past sales performance vs. the program incentives that these brand owners inflict upon us. We are bound to forecasting based on industry reporting compiled from past data.”. Hmmmmm.

Insofar as I could determine from steering the study towards these brand intermediaries, over 80 in all out of a total of 2200 survey participants that included brand owners that used classical and new age brand monitoring, there was a new opportunity to create a service that could support the mid-markets decision making process regarding which lines to carry, increase, add, drop, etc.

How many retail and wholesale distributor program incentives have to be co-integrated with the consumer’s  harvested perceptions and outcomes, as mined from public and private corpora? Well, the best known are cooperative advertising and inventory floor planing  carrying fees.  And there are more, depending on the industry. For example, the automotive industry is famous for incentivizing sales at the retail dealer level, with ties to specific models, lines, and trim options. One of the drivers for creating such a service for measuring customer perceptions of product performance and interaction outcomes, is that there is often a non-innocent motive behind a brand owner offering such incentives.

Now, these multi-line retailers are no dummy’s; they know that when they make an inventory decision, it will impact their bottom line. In motor sports multi-line, increasing Honda at the expense of Kawaski not only affects sales, but carrying costs at the floor plan financing level, especially if you guess wrong.

Providing a tool that affords a retailer or wholesale multi-line mid-market intermediary to work through incentives  co-integrated with extracted and graded consumer perceptions and interaction outcome metrics, will allow a more intelligent decision making process.

The challenge: create a great interface for the subscribers, create the statistical and stochastic engine for assigning value to the perceptions and outcomes as expressed in the corpus, and create an extraction methodology using the latest work on modular ontologies and symbolic substitution.

Those broad challenges are all in the wrong order of course, but they all have to be executed with aplomb. And I am working on it and pitching the more detailed plan of development strategy, market evangelism, and partnership issues that need to go into such a soup. So far, I have been all around the bay area, and several companies are taking a hard look.

Let’s see who pulls the trigger first – a big leviathan, or a scrappy startup.

One thought on “Brand Monitoring and Text Mining Saved by the New Science of Consumer Perceptions and Interaction Outcomes

Comments are closed.