The limits of algorithms in competitive intelligence?

Everyone who’s been to one of my courses in the past 587.6 years knows how adamant I am that analysts and managers who want to avoid being information “stick fetchers” must, absolutely must, include some recommendations- options for management to ponder- in their reports and presentations.

This is essential to getting intelligence work recognized. There is a legitimate role for librarians and archivists in the business world, but they are not looking to be recognized as competitive analysts. While they are proud to find answers to (somewhat random) questions, they are not in the business of impacting decisions and changing perspectives. CI professionals are.

There is a legitimate role for librarians and archivists in the business world...they are not in the business of impacting decisions and changing perspectives. CI professionals are.

When people with CI responsibility fail to understand how to properly form a recommendation – not as simple as one thinks- they can be easily outsourced to “research” in India or the Philippines. Then they become “senior consultants with 20-years’ experience.” And I feel for them; If only they listened to me.

Recommendations (opening specific options for management) have some rules to them, and Netflix of all enterprises teaches us a great lesson. 

Dystopian future

I like dystopian movies. It’s my cheerful nature looking at the rosy landscape of woke tyranny in this country that somehow propels me to favor dystopian, sci-fi type movies these days. And boy do we have these in vogue. Barren landscape, burning cars on bridges, and the collapse of civilization is one theme. Some viruses may be added.

Another theme is a bright future where totalitarian government with unlimited surveillance power (completely fictional of course will never happen here, OK, maybe in California only) controls every aspect of life and speech and communication. Behind the scenes or at the top there is always some corrupt power with malicious intentions. Cuomo, A/I, or the “System.” The hero, Heroine, or herobaby, fights back. Of course they eventually win.

I’ve watched maybe 13 of those in the past year and they keep coming. I am now dystopian-out. I can’t even look at another smoky bridge or Cuomo’s face. Why do I get them? Because Netflix’s algorithm sees my preference and ceases on it (nice play on words, eh?) and in the “recommended for you” section provides me with more of the same. But what if I enjoy a variety? What if I want to expand my horizon?

Netflix is not alone. On LinkedIn, if I click “like” on posts, after a while I see nothing of different perspectives. Algorithms tend to congregate us in groups of “insiders” and “outsiders.” Right with what’s right, Radical Left with Radical Left. Occasionally I’ll also get a promoted post such as the “Massage Envy South Florida is hiring massage therapists.” It’s refreshing, though I can’t help but wonder the marketing wisdom behind targeting me.

There is serious implication here for CI people making recommendations. Giving the intended audience (management) more of the same type of competitor warnings is consistent and persistent and may even get the message across, eventually. The advice of “don’t give up” is real. After all, I did watch 13 movies all clones of each other except for some speaking Czech.

At some point, however, you must change the algorithm. Management tires of the same crying “wolf”. They just stop listening to you. Risk mitigation is an important part of CI if done well, but it can’t be the only message. Opportunities are way more intriguing to management. In my book (literally and figuratively), every threat has an opportunity – two sides of the same coin of early warning based on changes in industry structure. Opportunities are much more difficult to spot and deliver but they are like a romantic comedy popping up on my Netflix “you may like this too” list- full of promise and pink hope and heartbreaking reunions.

At some point, however, you must change the algorithm. Management tires of the same crying “wolf”.

Netflix competitive strategy and CI “platforms”

If we are talking Netflix, it’s impossible not to be utterly impressed with its unequivocal victory over its main competitor in a crowded field. Netflix doesn’t only teach us about algorithm limitation. It teaches us amazing lesson in competitive strategy.

Have you noticed Amazon’s Prime Video streaming has lost the war with Netflix? I haven’t watched Prime is some time not because I suffer from Bezos Derangement Syndrome but because Prime has hardly anything new or worthy to watch. Maybe it’s because the Hollywood stars decided to wait for their fee to reach $15 an hour, or maybe Amazon has given up. Whatever is worthy, costs extra in “premium channels” which I refuse to buy on principle seeing as I already pay for Prime.

The victory of Netflix is so complete not because of my sense of injustice, but because of massively out producing Amazon. Netflix generates new originals at such speed, Amazon gets dizzy following, being busy with AWS and some marginal retail business. So, Amazon Prime Video just basically died. RIP.

There is a lesson in this to strategists: You don’t have to be truly innovative- Netflix original films are bad. Really bad. Not an original thought in sight except of producing in “exotic” locations/languages like Polish, Russian, Portuguese, Korean. Maybe it’s cheaper. Maybe it expands viewership (though most Americans hate reading subtitles). The point is you just overwhelm the competition. It’s a legitimate winning proposition under certain circumstances (of course, you need to understand the role of circumstances, and that’s why spotting opportunities is so much harder than mere data dumps.) Imagine Danone introducing 70 new flavors of yogurts a month. Nestlé will probably just give up.  

It’s a legitimate winning proposition under certain circumstances (of course, you need to understand the role of circumstances, and that’s why spotting opportunities is so much harder than mere data dumps.)

The opposite is true with the flood of so called “CI platforms.” Overwhelming the users with constant barrage of data, news, reports, more news, more data, eventually turns the users off the flood. And no matter how narrow or wide the algorithm searches and distributes, users I talk to just end up not paying attention at all. At the end, we are back to… “seasoned CI professional with 20 years’ experience with a retail company looking for opportunities.”

Alternative perspective: Algorithms can be overkill. Know their limits. Many (though not all) CI platform are disappointing for the same reason. Though it seems every week another pops up, they are not a panacea for the most important role of CI. Managers and analysts should try and expand their random “search” to include areas not typically related to their industry or products. Serendipity is a powerful source of insight in CI. Insights are rare and valuable and keep you employed.

Insights are rare and valuable and keep you employed.

I got my idea for the first war game methodology to be used in commercial enterprise in the US back in 1982 when I read Orson Card’s Ender books. If I used a CI platform it would have probably recommended War and Peace.

If you want to glimpse the deep difference between information practitioner (“research”, archive, demi-ci, competitor price tracking, platform addicts, data piling, stick fetchers) and intelligence analyst to decide if CIP? certification is for you, I will be offering a one-time-only, limited edition, personalized, customized by algorithms webinar on April 22 and 23. Seats are limited to the first 10,000 attendees. Hurry up! Your last chance for meaningful employment! https://academyci.com/17901-2/