Brad Ito: CTO, Retina AI
Have you heard that you need to map your Customer Journey? That all the real data-driven marketers are using a 360 Customer View?
These days there are so many technology offerings for marketers to use to manage their data that it can be dizzying to try to cut through all the hype. Based on over a decade making marketing data platforms, and working with clients on their data, this is a simple and actionable list of questions you can ask yourself to make sure your systems are up to par.
Most folks we talk to score at about half or fewer of these questions
- Is your data in one place?
- Is your data checked for validity?
- Can a marketer answer their own questions from the data?
- Do you have an identity resolution solution / strategy?
- Can you correct past data?
- Is your data available at the speed that you need it?
- Can you use per-customer data and segments in other systems?
- Can you enhance your data with advanced analytics?
The best part about this is that one can get an easy yes or no to each one of these. And you will immediately know where you measure up on the path to being fully data-driven. Most folks we know score about half yes-es or so, so don’t feel too bad if you don’t score perfectly.
1. Is your data in one place?
For those with one system for your marketing automation, one system for customer support, and top-of-funnel engagement occurring in Facebook and Google, it can be difficult to put the pieces together to really understand how customers experience your brand. Instead of trying to shoehorn data from one of these external platforms into the other (for example, trying to put all your offline events in the Google Analytics), own your data and the way in which it is stored. Using a data platform that is designed as a data platform will pay dividends with all its pre-built connectors, and features.
The original way that folks would centralize all of their data was in a CRM, but those have difficulty scaling to web-scale amounts of traffic, and were never designed to top-of-funnel events that occur before one has a customer’s email address or phone number. Enterprise Data Warehouses (EDW) and Data Management Platforms (DMP) do collect all of the data, but will make you jump through a bunch of hoops to make use of it. What folks will typically want is a Customer Data Platform (CDP)
2. Is your data checked for validity?
Your data is only as valuable as it is trustworthy. Tracking and technical errors can cause gaps in the data, duplicate data, mislabeled link tracking and more. You need to have some internal process for regularly sanity checking your data, and fixing or mitigating issues found.
Minimizing the human touchpoints to enable accurate and automatic tracking can be an important part of keeping data clean. But even fully automated marketing systems will have technical hiccups, and you don’t want to be caught with egg on your face for making an expensive decision based on bad data.
3. Can a marketer answer their own questions from the data?
Too many of the old “enterprise” systems force one to dig through obtuse user interfaces, or interface with the raw databases via technical SQL queries. That then turns into whole teams of analysts translating marketing business questions into the technical language of the data. Even with the best such teams, that added turnaround time means one asks fewer questions, and does less exploration of new ideas within the data.
Modern systems use flexible and powerful dashboards to expose all of the data for easy querying. It should not take a team of experts to pull up basic KPIs. And indeed there’s a wealth of research showing that we are at our most creative when able to explore questions independently.
4. Do you have an identity resolution solution / strategy?
All of your different customer touchpoints will have to track your customers using their own ids. From cookies and links to names and nicknames, from multiple email addresses and customer support calls to misspelled mailing addresses, resolving all of those touchpoints into what activities come from the same customers is hard.
Nobody has quite cracked this entire nut, and there’s vendor consortiums and new innovations happening in this space right now. The essential thing is to have a single unified solution or strategy. Without a unified approach being used across your platform, answers to simple questions like “how many customers do you have?” become impossibly complex and distracting.
5. Can you correct past data?
Mistakes happen, and you are going to need an ability to correct them. It’s important that your marketing data platform be able to represent the best and most current representation of real customer experiences possible.
Perhaps an email campaign went out with bad tracking links, you’ll want to correct that in your reporting. Maybe you only get some details about marketing spend in one channel after a 30 day delay, and you want to put in preliminary numbers that get corrected later. Or some channel partner only provide stats in the form of an Excel document. It’s important that your data platform support these sorts of edits/rewrites of history in a clean and simple manner.
6. Is your data available at the speed that you need it?
Customer data flows with different delays depending on where it is coming from and how it is being collected. There may be a continual stream of purchases and page views from your ecommerce store, but perhaps your advertising partner reports back performance at weekly intervals.
Take a look at all of your sources of data with time in mind, and ask yourself whether you are getting the right customer data in time to make the right decisions about it, and to take action when needed. Strategic decisions about the types of campaigns to run can happen more slowly and thoughtfully at a weekly level. Whereas cart-abandon marketing automation emails can happen on the next day. And on-page or in-app experiences may need to show different content or offers within less than a second.
Note that enabling faster flows of information will often require more investment in technologies to enable it, so also make sure that you invest for your own greatest needs.
7. Can you use per-customer data and segments in other systems?
All this data doesn’t do you any good unless you can do something with it. It should be easy and simple to come up with a new customer segment and deploy campaigns to them automatically. And 1-to-1 marketing requires that it be easy to pull up a complete view of all of the activities of your customer when communicating with them. Your data platform needs to unlock your data, not lock it.
Your data platform should have connectors for marketing automation, in-app pushes, chat, social media audiences, and whatever other channels you are using.
8. Can you enhance your data with advanced analytics?
There’s a new wave of advanced analytics helping to predict, model, and explain customer behavior, and you don’t want to be limited by your data platform. Whether it goes by the names of data science, machine learning, or artificial intelligence, these new analytics ingest large amounts of data, and output everything from per-customer lead scoring and churn prediction to predictive lifetime value and product recommendations. But if your platform can’t plug in to these you will be missing out, and letting your competition eat your lunch when they do leverage these tools.