Indeed, Gary it has been a long road to finally put fingers to keyboard to talk virtually about this topic. I am not sure you truly want me hijacking your blog for a while but away we go….release the Krakken!
Let me somewhat correct you in what I was finding with regards to personalization. A lot of this experience comes prior to EY but still holds true for the most part. The honest truth is most organizations aren’t doing a lot of personalization at all, nothing. Not even low hanging fruit like looking at previous purchase or content-consumption behavior to make informed decisions for the next best action/offer for these folks in personalized emails. Even with the insane amount of emails we all get on a daily basis they still are largely not personalized. With the web experience itself it is even worse.
A lot of big organizations are still trotting out giant monolithic offers or content to folks and hoping a percentage of folks like them enough to convert. And repeating this Sisyphean task over and over again. I still see very little usage of previous behavior or predictive behavior being used in seemingly simple things like email. In these cases, I’ve given these companies my permission to spam me (affectionate term) and yet they don’t capitalize. I mean we’ve been doing this for like 18 years now and we (the royal ‘we’ for Lebowski fans) still suck at email. It is a travesty.
I’ve sat with countless companies and heard them lament the fact that they have web data over here, 3rd party digital data over here, their CRM data somewhere else and their true customer analytics geeks somewhere else and nowhere did a Venn diagram form with an intersection with goodness. Like I said before, companies struggle mightily just getting this right for caveman stuff like email…now imagine trying to put people on Mars like personalizing the website in real-time to actually give the right offer to you.
Why is it so bad?
The roots to these problems are somewhat complex and different for different organizations. I’ll rattle off a few off some of the most common reasons and then I’ll actually answer your question (I didn’t forget).
A few of my favorite reasons I’ve heard why organizations don’t do personalization:
- Data siloes based on organizational setup in the company (ugh) – this is legit, real and scary
- We actually don’t have a way to connect our web behavior to our CRM – this is pretty universal
- We barely have time to create one creative let alone a B, C, D, E, etc – how in the world could we do personalized offers?
- The people that make decisions about what offers/content won’t let them try it or don’t believe in it
- Some organizations don’t have the backend to support or fulfill the personalized offers – which actually is a legit concern. It is one thing to throw something on the web but you have to be able to fulfill it.
OK, so first for those that are interested in doing personalization but are being held back for some reason listed above, you’re not alone. In fact, you are in good company. Everyone come in for a big hug. If you are interested in solving some of these I am sure Kelly on our team can help you (shameless plug).
Now onto your actual questions Gary…
I am going to try to answer it like this…companies universally aspire to give the right offer/content to the right customer at the right time. And somehow do it at scale. I think that is the heart of the matter. I’ve never met a company that varied from that. I believe the technology is all there to do this. It really is, but companies aren’t all at the right point in their own life cycle to do that. I think it is important for them to dream big but not lose sight of where they are on the journey. This is the jetpack we were all promised with digital marketing.
From a tools perspective…What I’ve seen is that companies are trying to do personalization by using testing tools to do the job. And really when I say ‘personalization’ I mean providing content/offers based on a set of rules that you’ve tripped over your past or current behavior. No real predictive scoring takes place or querying data from within the company walls (other than the occasional DMPs of the world) into account in the rules for decisioning. Basically, what you are getting is someone’s guess about what the group of folks should see. Example – I am a marketer and because you’ve viewed shoes on this visit (or if I am crafty, the last visit) I’ll recommend socks to you. It might be that the geeks have crunched the numbers and determined that people that view shoes love to buy socks (these are awesome socks). Maybe it is a predictor. But I don’t see companies doing stuff like that. I see companies making a gut feel kind of decisioning, such as “Well, when I buy a shirt I definitely buy pants” so everyone should. That stuff might be good but might be horribly wrong, I don’t see folks keeping score.
The second flaw is that maybe I am a megalodon-spender (new term I am copyrighting) on this site and you should be offering me like an Armani suit or something instead of socks as I am a high-roller. The problem is because you are only using my visit behavior and because your testing platform and CRM system aren’t friends, you make errors in judgment. Obviously, you can’t ‘know’ everything, but you get my drift. You could use this same thing with banks to grab credit scores or ad value on a media site, etc to keep me going.
Additionally, with most of the testing platforms or even the ones that do black-box self-learning, you have no decent way to open it up and plug in your own companies expertise. Here's some feedback I found fascinating. The most interesting quote was and I am paraphrasing “...if we did what the platform told us to we’d be like everyone else, we view our internal analysts as a competitive advantage and want to leverage that.” There are plenty of vendors that offer platforms for personalization that choose the best offer for you but I heard from multiple organizations that wanted the freedom to plug their own analytical models, their own customer knowledge from outside into the decisioning in order to choose the right offer/content for customers. From what I’ve seen in the testing platforms of today that is clumsy as hell.
Now…for the semi-counter punch to all that. I honestly think true real-time scoring decisioning is overrated unless you are in a select industries. I really do. I think you can do a lot with nightly models and reference the scores in conjunction with rules you have set that are grounded in behavioral analysis. Often that stuff probably doesn’t drastically change intra-day. So my thinking of today is the need of an open box to allow companies the freedom to do both…do simplistic “You viewed shoes, so you need socks!!!!’ decisions’ as well as complex decisions using analytical scores and CRM data that can be queried to be used to render a decision. Let’s call it hybrid…we need a closed box for some and some folks need to get under the hood to soup it up.
So, with all of said that I don’t think consulting has changed my opinion. I still think companies are still in the primitive stages but aspire to put a man on Mars. Goals are healthy.
And…did I answer your questions?
Comments