ArticleBest PracticesNovember 6, 2018
Messy Data? Join the Club Image

Messy Data? Join the Club

By Dan Steinman

So you’ve been wanting to get serious about managing your customers and using data to drive insights and actions. But your data is a mess, right? Is that the end of the story? It better not be.

Let’s get a couple of things out in the open:

  1. Everyone’s data is a mess
  2. We live in a data-driven world
  3. You can’t survive in Customer Success without quality data

Sounds like an untenable position, doesn’t it? I’m going to argue here that it’s not.

Customer Data Is the New Content

Some of you may remember when content started to become a big thing. I remember it well because it coincided with the growth of the internet as a marketing vehicle, but more importantly, with my time at Marketo when Marketing Automation (MA) was all the rage. Content was required at every turn. Lots of it. Many people threw up their hands and said they didn’t have enough content nor did they have enough people to generate content.

Guess what? The market did not care. The appetite for content was voracious, and those who figured out how to develop or curate it won and those who didn’t lost. It was as simple as that. Look at the quantity and quality of content that came out of Marketo and Hubspot as examples—and it was vital to their success as category creators. But it wasn’t critical only to category creators. Everyone had to get on board, and they did. Or they didn’t survive.

Today’s need to be data-driven is analogous to that content-driven world. It’s not an option. Those who do it will thrive. Those who don’t do it certainly won’t thrive and may not even survive. To go back to the Marketing Automation world for a moment, data was a key element in the success of companies using MA tools too. Who visited our website? Which pages did they visit? Did they fill out a form? What more do we know about them now if they did? Did they receive our emails? Did they open or click? And on and on.

Who Owns the Data?

The key difference between that world and our Customer Success world is that much of the necessary data in that world was created BY the tool. At least initially in Customer Success, most of the necessary data came from other systems, not from within. Therein lies the particular challenge. Therein also lies the excuse if you want it—the data you want is in other systems that you don’t have control over which means throwing your hands up might be a logical response. Logical? Perhaps. Optional? No.

Let me share another story with you from my Marketo days. I had hired four CSMs with the right combination of marketing, product, and customer-facing skills. They spent all day every day on the phone (or WebEx) with our customers. One day I was out on the floor listening to their conversations and I had a light bulb moment. It was abundantly clear to me that every single customer we spoke to was better off (healthier) at the end of that call than they were at the beginning. We never took customers backwards. But what hit me was a Management 101 realization: we were doing good things, but were we doing the right things? In other words, were we talking to the “right” customers? This really is task number one for any manager—making sure the most important things are the ones your team are spending the most time on.

Implicit in the above question is the need to determine which customers are the right customers. This won’t be the same for every company even if we speak in generalities. For example, should we prioritise at-risk customers or those who have the most growth opportunity? Not an easy question to answer.

For me, it was an easy question at that time. With so many customers and a churn rate higher than we could bear, our time was better spent finding customers who were on the path to churn and turning them around rather than making happy customers happier. The only data that I had easy access to about customers did not give me any insight into risk. It was valuable secondary data for assessing customer health (more on this later), but not actionable for my task at hand, which was reducing churn.

This leads us back to the data question. Without the right data at the right time, you are hamstrung just as I was. And unfortunately, the solution to this challenge is not easy. But who gets paid to do easy things? I sat at a conference of CEOs recently next to one I know quite well. This question of data integration and using data to make more intelligent decisions came up. One CEO stood up and talked about the complexity of their systems and the priorities of their IT people and how messy their data was and how hard this problem was, etc. My friend leaned over to me and said, “If his board is listening, he should get fired this afternoon.”

It’s a no-excuses world, especially for CEOs. But also for Customer Success leaders. I’ve often said that one of the key skills of a great CS leader is influence management. This is a perfect case-in-point. The resources that can help with this problem don’t work for you. But you better find a way to influence them or you’ll be updating your LinkedIn profile real soon.

5 Data Points You Already Have

I know your Salesforce (or CRM) system is in disarray. And the longer you’ve owned it, the worse it is. Join the club, because everyone thinks this (or else it’s actually true for everyone). In any case, I’ll argue that it’s not actually a complete mess. Seldom have I seen a CRM where the following information is not pretty darn clean:

  1. Original contract date
  2. Original contract value
  3. Successful renewals
  4. Closed/Won upsells
  5. Current contract value

This set of data actually provides a pretty good starting point for a customer health score, which is at the heart of doing data-driven Customer Success. Based on the above information, here’s what you know:

  1. How long they’ve been a customer
  2. How many renewals they’ve done
  3. How many upsells they’ve done
  4. How much their contract value has grown
  5. Annual growth of contract value

Nothing is a clearer indicator of customer loyalty than the spending of additional money on your solution. They can maximise the adoption of your product and give you tens on every NPS survey, but nothing speaks louder than a customer pulling out their chequebook (that’s how we used to spend money) and handing over more of their hard-earned cash. Of course, it’s possible your customers are loyal only because they feel they don’t have any choice. If that’s true for you, it won’t be for long. The market no longer allows it. This is the beauty of SaaS and subscriptions. More competition. That’s good for customers, but harder for us. Welcome to the Age of the Customer!

2 Data Points You Can Easily Get

Now if you add to that initial set of data a couple more things, you can establish a health score that, while not perfect (none ever are), is directionally accurate and pretty actionable. There are a couple more pieces of data that are typically well-tracked or are completely in your control:

  1. Support Case data: How many, what priority, how long they were open, what areas of the product they are associated with, etc.
  2. Surveys: How did they score you and what else did they say?

On the second one, if you aren’t doing surveys, it’s pretty easy to get started, and this is one set of data you can completely control if you own the process.

Now if there’s any way of also getting data about how customers are using or not using your product, you’ll be golden. This is most likely in a proprietary database (not a system), but it’s gettable. If it’s not, then today is the day to start asking for it loudly and often.

Your VP of Sales wouldn’t put up with not having data on your competition or critical information about your prospects such as size of company and industry. In the same way, you need to state your case for the data you need to run your business too. This is one of many reasons I encourage (demand?) that CEOs empower their VPs of Customer Success and give them ownership of the company retention number. With the ownership of that number comes power—the power to influence other orgs to support you in your otherwise-impossible mission to deliver to those high retention targets.

You Don't Need Perfect Data

Even without that usage data, you have enough to build a health score and start using it proactively to change or accelerate customer behavior. The data doesn’t need to be perfect, just directionally accurate. The same is true for your health score. By the way, if you apply the same health scoring methodology to all customers initially, the score itself may not be perfect but each score relative to all other scores will be perfect. If Customer A’s score is 35 and Customer B’s score is 71, I know that those are accurate relative to each other. And THAT is powerful.

And with a decent starting point, here’s what will happen: You will start taking action based on the data you have and you’ll start to see results. You’ll also find out that some of your assumptions about healthy or unhealthy customers were wrong. And that’s HUGE! What could be better than eliminating bad assumptions and narrowing in on the things that really matter? The last thing that will happen is the biggest. The use of data to take action and drive business results will cause a couple of ripple effects IF you don’t get deterred by them:

  1. The data will get better because people using it will see errors and anomalies and ask why. DO NOT let them stop using the data at this point. They will try because it’s easier to find things wrong and dismiss the whole idea than to work on it with you and have to exit their comfort zone and change their behavior. Keep them on track and relentlessly fix whatever they find.
  2. Users will ask for more. The need for data is unique in this way—the more you have, the more you want. It’s not unlike our internet surfing behavior. We often start a process by looking for a very specific piece of data. Occasionally that satisfies, but more often than not, it just leads us down multiple paths exploring other interesting articles and information. That’s why, after 20 minutes or an hour or longer, we often find ourselves asking, “What was I looking for again?” There’s always more information and we want it. Useful data almost always creates the need, or at least the request, for more.

Building a Data-Driven Culture

The use of data will improve the quality of that data and will also beget the need for more. And more data, with some logical limitations, will make you and your team better. If you provide the set of data I’ve suggested above to your team, I guarantee you they’ll start asking for more within a week.

  • “It’s great that we can see how often users log in, but I think it would be better to see how much time they spend in the app when they do log in.”
  • “It’s great to see support case information, but I’d also like to know if they are timely in paying their invoices.”
  • “It’s awesome that we’re getting NPS results from our customers, but why can’t we break that down between executives and end-users?”

Your challenges will not end, but they will be more positively directed. This is a much better place to be. Many have done this before you and with equally messy systems. The time is now to stop whinging about it and start doing something about it. It won’t fix itself. Start with something that is good and build on it. Or get someone from IT or your executive team on your side.

Use my line: “Why do you think Sales can’t do their job without relevant and valuable data but I can?” Or, if you think it would be helpful, have them contact me. I mean it. Have them contact me. I can tell them why and how some of the best companies in the world have attacked and overcome this challenge.

It's doable. And imperative. What are you waiting for?

Picture of Dan Steinman
Dan Steinman GM, Gainsight EMEA

Leave a reply

Your email address will not be published. Required fields are marked *

Shares