5 tips to address the scepticism when taking technology solution to market.

Or how to avoid the “this is BS” reaction

predictive analytics on AWS

Picture this. You are Sisyphus. The greek mythological figure that everyday pushes his boulder up the mountain and, as he reaches the top, sees the boulder falling back down. So that he has to start all over again the following day.

Now picture this. You are working for a company whose has a cool technology. Let’s even argue it is an AI based technology (could be anything but let’s pick a popular one for the sake of it). Good. You must be working with very smart people. And you work in sales? Nice, so you get to talk to a lot of people about this AI stuff. Exciting times.

Except that, you probably soon realised that there are some commonalities between your day job and Sisyphe. Indeed, every time you are meeting a prospect, you have the hard challenge to convince him or her that this great AI technology can sort out some of his / her problems. And over and over, once you explained what you do and how your customer use your product, you hear this comment: “This is BS”.  

Excuse me? This is what??  

What??! What is going on? What about this 2017 is the year of AI (which, if your sales and marketing focus on AI rather than on perennial business issues such as fraud, churn, efficiency of teams, etc..., begs the question of what will happen to your company in 2018 but let’s not go there). Why are prospects doubtful? How to address this?

Yes, despite the hype, selling technology based solution is difficult. At PredicSis, we are lucky as people can test our solution in one click (on here: https://demo.predicsis.ai). But let’s be honest, it’s not because it is AI “stuff” that people open their wallet and free up their diary just to see you. And there is a fair amount of skepticism out there.

There is a lot of BS. You have to lower the SB

Let’s be honest here. The reaction of a prospect is indeed justified. When it comes to AI, there is a lot of BS in the market place. And, let’s face it, AI has been around since 1956 in a workshop in the university of Dartmouth so we did not wait until 2017 for it to generate value for business. And at the moment, there is no shortage of companies jumping in the “AI” bandwagon. I am even aware of companies pushing the AI envelope to be riding the wave but, on the delivery side, have just outsourced some of the processes to people in low cost countries (true story). So the skepticism is justified. So, when talking to prospect, there are a few things I do when I want to address quickly the Scepticism Barrier (aka the SB) or avoid it to raise as much as possible.

5 steps to avoid your prospects turning sceptic

And I have to do it fast and concisely to make the most of my prospect’s attention. So, how I do I address this justified scepticism do I hear you say? Good question dear reader, here goes:


1- I rarely, very rarely mention the word AI with our prospects (admittedly a little bit more in the marketing literature. Guilty as charged!). I do a lot of prospecting and I very rarely mention the word AI. Fronting a dialog with “we’re an AI company” is, to any C-level prospect, a way to raise the scepticism barrier. Plus, frankly, who cares that it is AI, ML or anything else. What matters is the problems addressed, how they are addressed and what credibility your company has.

2- This brings me on a critical point for any sales process. I front the conversation by the problem we address. For instance, our customers are annoyed because they do traditional BI which tells them what their customers do in the past. But not what these customers are likely to do in the future. And company can’t act on past behaviours, only on future ones so why only have BI and not add a predictive BI brick to their stack. They are also annoyed, when doing BI, that it’s difficult to find correlation in their data so they pre-process data with PredicSis before injecting them in Tableau, Qlik, etc…  They are also frustrated that it takes them weeks or months to get predictive insights and would like to have these projects done in a few minutes and pushed in production in a couple of days. Or they are worried that doing predictive analytics is very costly both in people and technology so they are not progressing with it and are doing a lot of decision based on gut feeling (we resolved this this problem by enabling pay as you go model by using PredicSis via the AWS Marketplace for instance). These are some of the problems I highlight. Not that Predicsis core technology is ML, auto ML or anything else.

3- Credibility is important indeed. And a buyer is naturally risk averse. So I bring him some credentials very quickly in the conversation. Technology credentials. Predicsis is not the output of two guys in a garage. The core technology of Predicsis was first started back in 2004, developed by a large R&D lab. I bring quantified data that shows the technology was first developed when AI or ML was not all the rage. I bring quantified data that shows the tehcnology was first developed when AI or ML was not all the rage. I even send our advanced users to the page describing all the research that has gone into the product (you can read them here if you want but word of warning: heavy stuff)

4- I also bring customer credentials very early on to reinforce the credibility element. Admittedly, if you are a young start-up, with no big names that can be a struggle but there are other ways around this issue. So I highlight we work with very large companies like EDF, AMEX, Banque Postale Assurance (a major player in insurance sector in France). And I also stress that we work with tiny start-ups that have 10 people and need to enable their analysts without getting lost in algo or ML centric conversations. For instance, I send them the case study of how a BI analyst use predictive analytics and get results in minutes as shown here.

5- Last but not least, the proof is in the pudding. Once we’ve assessed there is indeed a problem costing our prospect time and money and that they are serious about fixing it, then we suggest they pull some of their data (csv files, even imperfect, suffice), run them via Predicsis.ai, surface insights. By moving onto the AWS marketplace, we have managed to remove the usual hurdles of testing PredicSis.ai. So this is a good acid test on how serious a customer is.

Nothing easy but following these steps, I find that I hear less dubious customers when I explain the problem we address. And, as I carry that boulder uphill, I don’t hear often prospects telling me: “this is BS”.

If you have other ways to address the scepticism, I’d love to hear your views on how you address the scepticism barrier. If you have “been there” and taken measures to address pushing that boulder uphill for little results, it would be great to hear some practical tips!

Intrigued? If you want to get a sense of PredicSis.ai, we made it easy for you through our Free Trial page