The CEO Manifesto: Capture Everything Once.Published: Wednesday, 12 July 2017 by Matthew Tod, Head of Analytics, D4t4 Solutions plc
In working with many companies over the past few years I am struck by how few have collected the data they need to power new digital, customer-centric or advanced analytics initiatives. Data production is always taken as a given and almost never receives the attention it needs; consequently, data is frequently the Achilles heel of many projects. No data, incomplete data, lack of permission, speed limitations, lack of depth, lack of access and confusion over ownership are just a few of the issues I regularly see hindering progress.
The reasons behind this failure are interesting and come from a lack of a joined-up data collection strategy. Traditionally data from digital channels is collected by tags and logs that generate the data needed for a specific application or vendor. This means that data is captured multiple times, and then used for a single application. Each silo acts alone, unaware of what other data is being collected and unaware of what is being done with the data. What is worse is that there is a cost to creating, testing, deploying and updating every tag or log. How big a problem is this? Well many organizations will have digital data dispersed between 15 to 25 different silos in my experience.
But there is no need for data to be the weakest link. The best organizations I have seen have adopted what I dub the CEO Manifesto - “Capture Everything Once” – and then use it in many, many applications, ranging from the simple to the most advanced AI-driven projects. It seems like nothing more than common sense in the digital age, yet most businesses don’t do it!
So why are these leading organizations taking an approach to Capture Everything Once and then using repeatedly? In talking with them I have found at least ten reasons as to why they are not following the crowd and instead forging their own path:
1. Simplification reduces confusion and saves time. If you have several watches it is unlikely you confidently know the time, and so it is with data. A single source of the truth simplifies everything.
2. Cost reduction. It’s very simple, but if you design out the requirement for a tag manager and reduce tags you have reduced testing and update costs. These costs are frequently significant, and as more digital channels appear costs scale in a linear manner.
3. Agility is dramatically improved if the data is already being collected; it is simply a matter of accessing it and then using it. No delays as tags are deployed and data is captured.
4. Compliance is much easier to enforce if there is a ‘single throat to choke’ and indeed this may be the only means to comply with the pending European GDPR legislation.
5. Control of the data means they can reprocess files at will and create new data feeds as required.
6. Connectedness between data uses provides for a rich data environment in which AI and ML can thrive. For example, if you run a survey and then want to see how the results are impacted by online experience, or if you want to append profiles with data tracking individuals’ interactions with specific campaigns, then the link will already exist if only one data collection tool is used.
7. Completeness means adding experiential data (things that change the experience but are not clicked on e.g. product reviews) and external data such as local crime stats to create an enriched, contextualized data set.
8. All channels are collected by a single tool for completeness and ease of connecting-up a customer journey across touch points for orchestration to both adtech and martech systems.
9. The right speed of collection and feed is important; some applications demand genuine speed (less than 0.25 of a second) where as other applications such as segmentation can happen at a much slower pace.
10. Customer profiles are at the core of almost every initiative, so creating these only once, automatically at scale and speed, is a critical part of the strategy leaders seem to choose.
Many organizations are not really in control of their data; they have sub-contracted collection to a vendor who holds and controls that data on their behalf. It is not real ownership in the sense that the owner can reprocess their data, extract it straightforwardly, share it between tools and easily take advantage of it in a myriad of different ways. To do that organizations have to decide to have Meaningful Ownership of their data, and take direct responsibility for capture, transformation and then sharing it with their chosen partners. In these cases, data is a genuine asset that can be leveraged for competitive advantage, rather than being something a vendor holds hostage.
Vendor independence is another important factor organizations must consider. Forrester talks about the need for “solutions [that] can integrate with multiple other digital analytics and optimization technologies” ¹, and with the need to orchestrate the data easily across multiple platforms today, self-build, single vendor and rip-and-replace strategies are simply not an effective option. Better instead to use data capture software that can be easily integrated with your incumbent systems; that way you’ll maximize investment on both the data and your existing martech investments.
How do they achieve CEO?
I see this CEO approach in talking to our clients who have deployed the Celebrus Real-Time Customer Data Platform which is engineered to enable this vision. The data Celebrus collects is like the electricity that powers multiple devices in your home. It is a single, standardized source, available in the right format, at the right place at the right time. It might not be glamorous, but to those who understand data, it is a very precious enabler of their dreams.
¹ Source: “Vendor Landscape: Digital Intelligence Technology Providers You Should Care About”, Forrester Research Inc, February 17, 2017
LinkedIn: Matthew Tod