In a Harvard Business Review study released earlier this year, 99% of respondents reported that their companies are moving in a data-driven direction, yet only one third believe that they have succeeded at doing so.
It's pretty clear that being data-driven -- or collecting and using data to inform your core business decision -- is a good idea. But many companies and teams struggle to actually build a data-driven culture.
I started my career at Capital One, a company that epitomizes the concept of data-driven decision making. And here at LawnStarter, if there's one thing I can say we are, it's data-driven.
Here is my perspective, based on my experience, on how to build a data-driven company or customer success team. I should point out that this is not the way, but one way that I've seen teams be successful. After all, there's more than one way to skin a cat.
How to Build a Data-Driven Customer Success Team
1. Build data analytics into your culture.
In my experience, a data-driven culture is not something that can be slapped on. It must be ingrained at the cultural level. Lack of data-driven decision making can't be a quarter-long project one solves. And it's not as simple as hiring analysts and building dashboards. It needs to be in an organization's DNA.
In my experience, building a data-driven team consists of two aspects: hiring data-driven people and empowering them to use data effectively.
2. Hire data-driven people.
It's much easier to hire people with a certain mindset than it is to change one's behavior. Therefore, I believe that it's important to hire people that value data.
At LawnStarter, we went as far as to make ‘Be data-driven' one of our core values, and we don't hire anyone unless we believe they fit this profile.
Here are a couple tips on hiring data-driven people:
Recruit from data-driven companies.
One easy way to ensure your company is full of data-driven people is to go out of your way to recruit from companies that are known to be data-driven.
At LawnStarter, we've recruited people from companies like Mckinsey, Google, Capital One, and Indeed, all of which are known to be extremely data-driven.
Those who come from successful, data-driven companies have seen first-hand the importance of data and are naturally oriented to embracing it. Additionally, they will bring best practices and skills with them.
This is not to say exclude someone just because they don't have one of these companies on their resume.
Screen for data-driven employees.
It's important to note, hiring data-driven people does not mean you should only hire analysts, data scientists, and data engineers. It's very unlikely that a customer success rep needs to know Python or SQL, for example.
However, every person in your company should embrace and buy into the concept of using data to make decisions. It means letting go of one's ego when a belief is disproved by data. It means not taking shortcuts on ticket flows to ensure proper data collection. It means understanding one's own metrics and how to move them. It means testing new ideas vs guessing.
How to screen for data-driven people? Ask how they solve problems, or how they think they should measure success. Look for answers that show a thought process involving data.
3. Empower your team to leverage data.
Hiring data-driven people is not the only thing a company needs to do to be data-driven. A company needs to ensure that once data-driven people are onboard, they are empowered to use data to make decisions. If you bring on folks that want to use data in their daily work lives, but they struggle to do so, they're going to quit.
Here are some tips on how to empower your people to leverage data. Note, this is not a recipe, but rather a few lessons shared by me and others who have worked to build data-driven companies.
4. Make your data accessible and in one place.
In order for data to drive your business, it needs to be accessible. Seems obvious, but it's easier said than done.
First and foremost, this means connecting all of your data together in one place. Typically, this has required investing in a data warehouse, but these days, you can do this entirely with out of the box tools. For example, at LawnStarter we pipe all of our data into Redshift using only three tools: Segment to track events on our website, SerpDB to track our search engine rankings, and Fivetran to extract data from our CRM, help desk, and production databases.
There are different ways to measure things like active users, churn rate, and even revenue. Make sure you define these once, and use the same definition across teams. Otherwise, you end up feuding over which metric is correct, and create silos.
Fact tables are when you take a complex dataset, and transform it into a simpler set of tables that can be easily used and accessed. Typically you do this for a few reasons. One, it helps create the aforementioned single source of truth. Additionally, it makes it easier for those without advanced data chops to still use data.
Focus meetings around data, but give participants time to prepare and pre-read.
Whether it's a regular metrics review meeting or a meeting to share the results of a deep dive, it's important to give participants the chance to pre-read.
"We've done our best to build a data-driven culture at Break the Web, simply because backing up claims with data means better results for our clients," says Jason Berkowitz, CEO of Break The Web -- a Manhattan-based creative marketing agency. He adds, "Our weekly meetings require everyone to not only prepare summaries, but for everyone to pre-read and ask questions the day before to ensure the meeting is most productive."
Often there are questions whose answer exists in the data, but that will take time to uncover. If these questions are asked in the meeting, it is easy to have the meeting degrade into debate. If the big questions are asked ahead of time, those answers can be brought to the meeting, keeping all discussion grounded on the data.
Understand that data will break, build a process to handle it.
Nobody expects any technology product to be without bugs from time to time, yet many people expect data to be perfect.
Whether it's because of a human error, a change in how data is collected, or a technical issue, you can count on one thing: there will always be bugs. It's important to understand and embrace this, and create processes to constantly report bugs and maintain data sources. Otherwise, you risk falling into what Brian Balfour calls the Data Wheel of Death.
Teach and evangelize data competency.
Helping team members use data better helps everyone be more effective at their job. Additionally, it shows people first hand how impactful data can be.
Jackson Berry is an analytics consultant at Tableau Help, and previously spent three years working for Tableau's consulting division. He says, "I've consulted with several dozens of companies, and the ones that really succeed tend to make investments in training. Everyone wants to be data-driven, but the companies that succeed focus on facilitating better data literacy for each and every team member throughout the organization."
5. Measure individual, team, and company performance by metrics.
Few would disagree that in business, metrics or KPIs are necessary for measuring the productivity and performance of a company. Clearly defined KPIs make it very clear whether a team is performing or not. Or do they?
As it turns out, setting metrics and KPIs is easier said than done. In customer success, it's quite easy to measure customer satisfaction score and tickets solved per hour on a team and individual basis. But what about a role like content marketing, where there is a whole lot of variance from week to week. A blog's traffic can fluctuate wildly every month because of press activity, search engine updates, or because one post just happened to go viral.
Stefan Dubois, CEO of Survey Anyplace, recommends the following: "Next to result metrics, establish activity metrics -- these are metrics of activities that lead to results, and it relates much more with the day to day activities of the team. For example 'number of social media posts per month' could be an activity metric for the marketing team."
Being data-driven must be a cultural norm -- it cannot be slapped on like lipstick on a pig. In my experience, it starts by hiring folks who share a respect for data-driven decision making. And then, it's about making sure your team is empowered in using data to make decisions.