OK, weird title – but stay with me, promise it gets better and it is a great point on data use.
In the 1995 mini-series The Langoliers (here it is, in YouTube, in all it’s three+ hours of wonderful made-for-TV acting) based on the Stephen King book Four Past Midnight, we follow a plane that time-travels. Due to a glitch in the time-space fabric, passengers who were asleep as the plane crossed the glitch (including a co-pilot, else — who would land the plane?) end up in the past. Turns out this is not so good, due to limited resources our past times are not preserved as we would expect in traditional time travel. An entity, the Langoliers, “recycles” resources by eating the past briefly after it occurs and uses those resources to create a yet-to-be-experienced future of a few minutes ahead – think of an alligator chomping at the fabric of time-and-space constantly to get a picture of what the mini-series depicts.
Essentially, as the story unfolds, our lives happen in a +/- 10 minutes continuum that are constantly re-invented (yeah, too heavy – but that’s not the related part of this story). We simply move through these continuum as characters in a play, and the “stages” are recast for each scene constantly.
(OK, the basics are set, now comes the comparison to customer 360 – and the future of customer 360 – briefly stated.)
During Dreamforce 2018 Salesforce announced the rebirth of Customer 360. In case you were not with us for the past 30 years, customer 360 (the concept, not the brand or trademark – C360 for short) was the original promise of CRM. We will collect enough data about our customers so we can create 360-degree profiles (read, complete from any side you choose to look at them) and use those profiles to provide for our customers. Salesforce is using parts of Mulesoft and other recent enhancements to their platform to do this.
This is what CRM was supposed to accomplish, and was heavily marketed until about 10 years ago. The era of the “Big Data Myth” pretty much destroyed the concept of C360 by creating more data that it can be sensibly processed and used.
When Dreamforce kicked off I was in Wisconsin presenting to a wonderful crowd at the UWBEC Conference (in its 20th year now) and I was awesome as always (there’s a video, not sure I can share it – will change this link if I can). I was flying during the keynote, and my connectivity was spotty at best – I know, they try – so I decided I would let the keynote happen and catch a rerun later. As I landed, and not being a citizen of Twitter any longer (a year and 2 weeks as of this writing, I feel much better now – thanks), I was parsing through my text messages and emails to get an idea of what I missed during the almost 4 hours flights, when I noticed C360 was back in the headlines.
My first thought? I time traveled back a dozen years or so and we were back in Langoliers-land.
(phew, that was a hard connection to explain – but it gets better… there is a good angle for this ).
Now, given the history of C360, the Langoliers connection is a very good one, if I may say so myself.
The problem with C360 in the past, and the reason it failed, was that an organization wanted to “own” all the data necessary to create these profiles. In the process we invented weird ways to store and manage data: we had the data warehouse (and its cousins, data cubes), followed by business intelligence, data lakes, data swamps and my current
hated preferred model: customer data platforms (incidentally, I recently wrote a 2-part article on ZDNet about this, first talking about data usage, and then talking about data platforms).
All these are good models, don’t get me wrong – far be it from me to tell you that something is a band-aid brought about by lazy people unless we are talking about private and hybrid cloud – and they work for different companies to integrate data from different company-owned databases into a single repository (an evolution on the concept of ODS – operational data storage – for the current cloud-based world if you prefer), but they don’t accomplish the intended goal of collecting or aggregating all necessary and timely information about a customer. They simple collect as much as possible with the use for it “to be determined” later. Ugh. I could write two paragraphs about why this is wrong (or two articles, read above, or a 20-page whitepaper for Radius – registration required), but let me tell you that in five easy steps:
- Customers have data that is not captured by the company, but exists somewhere and is accessible (hello, Google)
- Customers have a sense of privacy (very little for some, including Alexa “parents” and social media over-sharers) and don’t share all their data
- Company does not understand that access, not ownership, is the only way to use this data on a timely basis (this is critical, and wish it was more widespread)
- Company does not get yet that customers want outcomes, not offers and coupons (for the most part, there is always a few exceptions)
- In the time it has taken organizations to understand this, the Langoliers came by and recycled the potential value of the data into a different reality
What’s an organization to do about this? Are we doomed to perpetually have over-bloated data stores about customers and incomplete insights about them due to lack of processing time and power to resolve interactions in real-time? Are we jinxed to constantly present customers with not-the-greatest-offers that don’t reflect everything we should know about them but we failed to collect into our data stores on time for processing in real-time?
Enter the new C360.
Now, Salesforce promised to make Customer 360 happen in the next year or so, I think GA is expected in June of 2019 – subject to change as per their Safe Harbor statement. I had early discussions on direction and positioning, but nothing I say here is final or more than my opinion. My opinion on this is almost my desire for them to use the time between now and then and take my desires and make them into messaging / action for their product (this is why I am called an “influencer” – I make things happen by influencing thoughts for other people and public supplication / shaming is part of it sometimes). If history is any indication, I will be somewhat disappointed – but a boy can dream, can’t he? Besides, any advancement is better than none and Salesforce is the closest we have today to make the new model work (at least until Microsoft realizes what they have and step in full-blown ready for battle)
The new C360 should be about timely use of data, not about collecting and aggregating into data stores. I have been talking about this for a long time, but my advocacy is to use data in real time and save the results, discard the data (you can save the calculations, meaning which variables went into making what decision by the analytics program, but the data itself is useless – with the exception of few places where you need to save it for compliance). The new C360 is about the calculation, not the complex mode of correlating data and knowing where to get what, where to store, use, and store again after calculation.
After all what interest you, as a business user, more – an outcome or the way you got to that outcome?
Gee, even the new-fanged common-core math model my kids use these days is about results, not calculations to get there. This new world of speed in which we live is about outcomes, not how to get to the outcomes – that’s access, not ownership.
The new C360 is about how to optimize the use of AI (aha – now I got your interest): build ever improving algorithms, compute quickly the likelihood of something happening, record the result and the outcome, analyze, learn, and repeat (automating as you go along). This eventually leads to AI in the organization, and in the process covers advanced analytics, predictive and prescriptive, and early onset of machine learning.
The new C360 s about using data, not owning, and certainly not storing. This is where all vendors failed before: tried to focus on what they don’t do best – store data in complex data models. As my friend Jesus Hoyos (one of the foremost implementation people I know for #MarTech and leading the digital revolution in Latin America) will tell you – data models are where good intentions go to die – there is no way to connect two (usually more) data models into something useful that can be used in real time, there is a way to use the data from those data models into a real-time computation and use the results of it – continuously.
And that, my friends, is how you avoid the Langoliers of data: you don’t live in the past, where they are eating away resources to recycle them, but in the future – where the possibilities are endless. And you need a platform to do it, make it happen, to be able to compute real-time using data from myriad places – not a storage place where data languishes until the Langoliers eat it away.
What do you think?
(disclosures: Microsoft is a client, Salesforce is a client, Stephen King is not my cup-of-tea as a writer, I hate horror books and movies and stuff, and I was always intrigued by the concept of time travel – which is why I watched the mini-series. I heard there is also a movie about it, have not seen it. All ideas herein, including omissions and errors, are mine as you’d expect and safe harbor says you shouldn’t make your decisions based on what I say may happen (even if I am right most of the time). No one can make me write something I don’t want to write, and nothing is intended by putting Salesforce name in this post, I am submitting a more detailed report on my thoughts to them through the year as we work together — this is public posting of an idea on how to re-do C360. if someone has the trademark on Customer 360, my apologies and fair use and all that – if no one does, well – there’s an opportunity for someone. And Jesus Hoyos is a friend of mine. Use of crossed out text is courtesy of Jon Reed over at Diginomica, a master at the subconscious influence of crossed-out text)