As latin people used to say, “nomen omen” (the name is a presage). A field so crowded that the Martech 5000 census (courtesy of Chiefmartec.com) has now grown past 7,000 companies worldwide selling software for marketing. Isn’t that mad enough?
As someone who has worked in marketing for more than 20 years — both as a client and as a consultant — I know for a fact that people in marketing are not very happy to learn another software. This is why usability at large is one of the most important aspects for MadTech SaaS companies, and yet I have noticed that many solutions in this field have been designed by engineers for engineers, not for marketers. The vast majority of these platforms (I am speaking out of a long, personal and painful experience with many of them) have been developed as tools, albeit sometimes very powerful ones. But marketers are already surrounded by tools and information, and yet 95% of their time today is spent doing the following:
- getting in and out of pointless meetings;
- preparing just one more presentation for some executive’s visit;
- executing a long list of tactical tasks that grows bigger by the day;
- trying to explain people that marketing requires knowledge and experience and no, not everybody is entitled to say if they like or dislike the last campaign;
- trying to remember why they chose this job in the first place (“Choose marketing, you will have fun and travel the world”, they said).
Only 5% of marketer’s time (I’m overly optimistic here) is spent doing actual marketing, i.e. thinking about the Brand and business strategy that is needed to succeed. So no, marketers DO NOT have the time to learn just another damn tool, nor they have time to use it. What marketer need is actionable insights, not mere data points to interpret. They don’t have time to connect the dots; they would like to look at the whole picture instead, and use that precious remaining 5% of their time to lift their head off their desks and cubicles, and think.
"Marketers DO NOT have the time to learn just another damn tool, nor they have time to use it. What marketer need is actionable insights, not mere data points to interpret."
And yet this is not the way MadTech solutions are designed. The purpose of these solutions, and of Big Data at large, is to empower marketers. But empowering doesn’t mean providing an infinite canvas where people have to endlessly meander before finding a path. Empowering means providing absolute clarity about the battlefield, so that marketers can focus on what really matters, the winning strategy. This is what separates good MadTech solutions from fancy but time-consuming bloatware. So let’s identify just three criteria to choose your MadTech solution wisely before the damage is done. Though I’m focusing on MadTech Data-as-a-Service solutions now, the considerations discussed here apply to every other field in the space.
It’s pretty obvious; spending even a second of your precious time on something outcome unreliable is absolutely pointless and dangerous. And yet the vast majority of Big Data MadTech SaaS have huge, and I mean HUGE, problems of statistical validity. We have already discussed it at length in this article about live digital data sampling. So before choosing a solution, take some time to investigate the methodology, the data sources, and how keen your vendor is on statistical validity. If their answer is a mere “we process A LOT of data” or even worse “we use very large samples of Big Data” well, run away as fast as you can.
"If your Martech vendor's answer on statistical validity is a mere “we process A LOT of data” or even worse “we use very large samples of Big Data” well, run away as fast as you can."
2. Insights, Not Data Points
A data point is an information. We could describe it as the manifestation of an underlying phenomenon — WHAT is happening, so to speak. “Consumers are buying more of Brand X this year, especially on channel Y”, is the description of 2 interesting data points. You definitely want to know about it, but then the question becomes “Ok, what now???”. To move forward you need an insight, i.e. the underlying reason WHY , the cause behind that phenomenon. The good news is that today there’s plenty of information to provide these insights, but it requires a powerful mix of best-in-class Artificial Intelligence and human analytics capability, which cuts the list of MarTech vendors by 70%, to say the least. Everyone claims to work with Artificial Intelligence, but that's not necessarily true; the definition of A.I. is inconveniently vast.
In the picture above we are showing the case about a ORIGINS product in the Face Cleansing category. ORIGINS is a popular Brand from Estée Lauder, known for its attention to the quality of its ingredients, and yet its Net Opinion Score (NOS) on the INGREDIENTS topic is very low. This is a data point; we now need to know WHY people are complaining about the ingredients.
"Everyone claims to work with Artificial Intelligence, but that's not necessarily true; the definition of A.I. is inconveniently vast."
By diving deeper, the system automatically extract people’s comments about the topic, revealing that consumers lack the necessary knowledge to read the ingredient list properly. Sodium Chloride (a.k.a. table salt) and Myristic Acid (a.k.a. coconut oil) look like two dangerous chemicals to many unexperienced consumers (our insight). Time to change the language on the ingredient list and educate consumers about ingredients quality with a focused content strategy (the Action needed).
3. Easy To Use
No Madtech solution is really of any use, if it takes a master degree, endless days of work and infinite keyword searches to get to an actionable insight. This means that the User Experience (UX) needs to be designed looking closely at the insight extraction process, to make it as easy as possible to get to the desired insight. This is not just to spare time; bad design could actually hinder the very possibility to extract a meaningful insight from valid data.
"No Madtech solution is really of any use, if it takes a master degree, endless days of work and infinite keyword searches to get to an actionable insight."
Here’s an example: a marketer suspects that consumers do not like the smell of the body cream she sells. She wants to investigate the topic on her Voice-of-the-Customer software of choice. First thing she has to do is to put together a large set of keywords that cover the “Fragrance” topic. Fact is that there are a thousand different way to describe a smell (typos excluded), and consumers can be veeery creative. After a lot of hard work and thinking, she will end up with a very limited list of keywords. She will then have to submit the set of keywords to the platform (not an easy task usually, as the programs usually use a classic search field), in association with the Brand name, or at least in a certain proximity. After all these hard work she might have an answer. Or maybe not, because the smell could be a problem every Brand in the category has. Or it might be something consumers simply complain about, but that does not influence final purchase decision.
Sounds familiar? This is because the extraction technology has been designed around keywords, not topics. Yes, of course topic extraction is a complex Artificial Intelligence business, but we’re in 2020, for goodness’ sake. Even if it is rocket science, Elon Musk has managed to land 2 rockets in perfect sync about two years ago.
As of 2020 there are 7,000 MadTech companies in the world, and growing. One reason more to be selective when choosing the solution you might end up staring at for endless nights.