How to Build an Analytics Team
Digital analysts are different from other people in the company. We always talk about digital analytics in terms of people, processes and technology. This article covers the ‘people’ part and why nurturing digital analysts is so important when learning how to build an analytics team.
You can read this wearing three different hats; a manager of a team, a team member, or a team member who wants to be a manager. Everything we cover here will apply to you.
This is a write-up of the talk given by Jim Sterne at GA4ward MKIII. You can find his slides here, and the recording is below:
We’ll be talking about ten management quick hits – let’s jump in with the first.
#1 – Manage up
Your job, when building an analytics team, is to protect the team from the organization. Because they are unique, they need special care. Analysts are seen as second-class citizens, they don’t break records and they don’t make decisions. They are dispatchers, not firefighters.
Your job is to run interference and provide support for digital analysts. You need to continuously explain and market why they are valuable. You need to educate your internal or external clients about how digital analytics tools inform decision-making.
If you’re not a data-driven company people, often in senior roles, will be hesitant to receive data. They may even view it as a judgment. You need to explain to them over and over again that you’re there to help them.
If you tell them something they don’t want to know, they’ll give you lots of reasons why you’re wrong. These might include any of the following:
- Fear of Data – I’m bad at math
- Distrust of Data – I don’t believe you
- Lack of Data Governance – My data is better
- Lack of Data Vernacular – That’s not what we call it
- Lack of Data Exposure – Crosstabs? Pivot Tables?
- Lack of Data Decisioning Process – Intuition got me this far
A client may ask to see numbers. Instead of saying ‘Why do you want to see them?’. Say ‘What’s important to you about this?’, or ‘What challenges are you facing?’. When you get to know them well enough, explain that the numbers will help them earn their bonus.
#2 – You should not be the smartest one in the room
If you are the smartest person in the room, you’re in the wrong room.
This is one of the biggest difficulties of transitioning from an individual contributor to a manager. The contributor is the person who really knows the code. They know how to put together a SQL query in the most beautiful way. That’s not you if you’re the manager.
Your job is to let them be smart. You should be worrying about staff meetings and administrivia. They should be coding, building a pipeline, doing analysis, and creating visualizations.
These skills don’t belong to one person. You help them build skills and get them to be the best at their ‘thing’.
#3 – Do not micromanage
We’ve talked to an uncountable number of digital analysts over the years. We always ask ‘What advice do you have for your old managers?’. The answer is always the same – ‘Get out of my way and let me do my work’.
Knowing how to build an analytics team you, as a manager, need to be comfortable with the fact that they won’t always get it right. They won’t do it your way. If their way is different than your way, it doesn’t mean it’s bad. You have to let them get on with their work – they might surprise you and do things you haven’t thought of.
If they don’t do it wrong, they won’t learn. After all, the only way to improve digital analysts’ performance is if they make mistakes. Let them do their own research and give them problems to solve, not tasks to do.
When someone asks you a question such as ‘How do I do this?’, the best response is ‘I don’t know, what do you think?’. Make them think, make them work on it.
#4 – Juggle personalities
When it comes to knowing how to build an analytics team, there are two types of personalities that you’ll need to deal with. These are the extroverts and the introverts.
The extrovert is someone who thinks out loud. Give them agency and encourage enthusiasm. You need to let them pontificate for a while, as that’s how they understand what they are thinking. They have to talk it out to see what they think.
Introverts are the opposite of this. They think a great deal before they start speaking. That’s because their brain is so full. There’s so much going on in their heads that they don’t know what to say first.
Unlike extroverts, who gain ideas when they are in a community with others, introverts need time alone.
Don’t think one is good and one is bad. They are just different. You’ll be dealing with the lone wolf (the coder or the modeller), and the teacher (someone who loves explaining). A leader is someone that is always encouraging others and who people go to for help.
You will succeed by encouraging diversity. Everyone has a different lens through which they view the world. The more diverse the lenses are, the more likely it is that you’ll come up with better results.
Remember, you need to defend the socially inept members of your team. Take them aside and do some soft skills training. Just because they aren’t the life of the party doesn’t mean they don’t add value.
#5 – Design individual career paths
The first thing to understand when building an analytics team is that everybody leaves. Nobody is going to work for you forever. They’ll get another job and move on to pastures new. The only solution is continuous recruitment. You always need to have people in the pipeline that you would hire when there’s an opportunity.
Experience is important, but you also need people that have the curiosity to grow. You need to create continuous learning opportunities (eg. the ability to go to a conference, a Measurecamp, etc.). Learning is essential if you want to retain digital analysts. If you don’t encourage people to learn, they’ll get bored on move on.
When you create a career track for somebody, ask yourself some questions. Are they interested in being on the path of the sole contributor? Or, do they prefer the management path?
Consider individual goals
People have different goals at different times of their careers. You should revise these goals quarterly to track changes. Ask questions like ‘What do you like most about your job?’, ‘What do you like least about your job?’, ‘What can we get you to do more or less of?’.
If you have a big enough team (more than 8 people) give serious thought to rotating tasks. For example, you might move someone from email analytics to social analytics. Social analytics is then moved over to search analytics, which is moved over to web analytics, and so on.
Encourage continuous growth
Not only do you want your employees cross-trained so that if an employee leaves, there’s somebody to back them up. But also to support continuous learning.
Remember, continuous learning means more than sending someone to a class. The best way to learn something is to give a class. So yes, send them to conferences, but also ask them to be presenters. It can also be a good idea to encourage employees to present to each other.
For learning, provide space and time for deep work. Large companies are starting to pick a specific day of the week when there are no meetings. Instead, they set a day aside for deep work.
#6 – Promote leaders, reward achievers, and replace cheaters
This is one of the key elements of knowing how to build an analytics team and managing digital analysts.
The employees that people go to ask questions to are usually prime candidates for promotion. You need to groom them to take your job. Ask them to take on tasks that are a little more managerial. In other words, give them the ability to step into your role so that you can take your boss’s job.
Rewarding achievers is common sense, they deserve it. They’re saying what they’re going to do and hitting their goals.
Finally, you need to replace cheaters. These are the ones that aren’t putting in the full effort. The materials they deliver have to be corrected all the time. To put it simply, they’re not quite on it. Of course, they could be A players in different positions or at a different company.
Letting them go is giving them opportunities to be great elsewhere. If you have C players on your team, you owe it to them to let them go. If you have B players on your team, you should help them become A players.
With that being said, don’t make them managers. B players hire C players. In other words, people that won’t replace them.
#7 – Culture eats strategy for lunch – Peter Drucker
Culture is not the motivational poster on the wall. Culture is what you tolerate. Is it okay to be late for meetings? Is it okay for people to snipe at each other behind their backs? If so, that’s your culture. It’s what people come to expect when they work for you.
Culture is the values that are expressed through rituals and practices. When building your analytics team, be very transparent about expectations, progress, and how things are going in the company.
Be transparent. For instance, if everyone knows the financials of your small company, they’ll understand what profitability means. They’ll know why the whole team needs to work for profitability.
In the absence of information, people make things up. Perhaps people believe that more products were sold this year than last year. Employees might think the owner of the company must be going on vacation and buying a yacht. If they had seen the company’s overheads, they might have a very different picture. By being transparent, you get to control the narrative.
#8 – Praise in public, criticize in private
Praising in public and criticizing in private is a good practice, except when it comes to cultural issues. These problems need to be called out immediately.
If during a presentation, someone shouts out ‘Well that’ll never fly!’ you call them on it straight away. That’s how people learn in real-time.
Other than that, thank people when they do something good, in front of everyone else. This is especially the case if there’s a member of the team that needs a confidence boost. Figure out a way they can hear you praising them to somebody else.
Your wins are theirs – if they did something great, that’s their achievement. If they do something wrong, that’s on you.
#9 – Solicit feedback
Soliciting feedback isn’t always easy. It can be even harder for your employees to give you an honest response. So, what can you do? Get somebody else to ask them open-ended questions about how they’re doing.
People are nervous to tell their boss what’s wrong, so they don’t. They don’t want to rat out their co-workers, so that makes them uncomfortable.
But if the third party says ‘I’m interested in your opinion. I’ve got some open-ended questions, I’ll take them and rewrite them,’, it can help put minds at rest. People that are working in a small team can tell who’s said what. A third party helps to avoid this issue.
#10 – Work on your eulogy
As we all know, at the beginning of Covid, there was a lot of uncertainty. People were asking questions such as ‘What do we do? And how can we help people? The simple answer was that this was not the time to be thinking about quarterly reports.
Instead, it was the time to worry about fellow workers. Nobody ever got up at a funeral and said ‘They met their quarterly goals’. They would say ‘They really helped others,’ or ‘They were always there for me’. Those are the traits that matter in management.
Letting people know that you care about them is a big deal.
#Bonus point 1 – Staff satisfaction
Soliciting feedback is about you. Staff satisfaction is about them. We’ve come across the best tool so far to help you with these differences.
This is the Q12 employee engagement survey by Gallup. It asks questions about the individual. These kinds of questions are marvellous for finding out if people feel they belong.
Below are a couple of other resources that might be useful to you.
- The digital analytics association has a competency framework, job descriptions, self-assessments, etc.
- A great book called Building Analytics Teams.
- An amazing 100-tweet Twitter thread about managing.
- Vivo Team and Core Sciences both have reports about how people work together emotionally.
Team learning versus team efficiency
Commercial organizations working with and for clients, look at team efficiency. How do you view team learning (investment) compared to team efficiency (benefit) with learning on the job as the default?
People aren’t machines. They don’t work at high efficiency, they’re always learning on the job. If the job was the same it would be automated. We are learning about the client, the marketplace, the process, and more. Learning on the job is the default.
As a manager, your job is to look at a project and scope it so that you have learning time built in. If it’s a 10-hour job quote it as 15 hours because there are factors you can’t see yet.
What should a business do that has a few A+ achievers carrying all the burden of productivity? How can you nurture these employees? What’s the long-term outcome for the business if it stays this way?
If you have a few A+ achievers carrying all the burden, the business will carry on as it is. But eventually, they will burn out and go somewhere they can work with other A+ players. You either raise the C and B players, or you get rid of them.
#Bonus point 2 – Generative AI
We had digital transformation which was about data. Then we had to deal with covid, which turned up the heat on this. Now Generative AI will turn everything on its head.
Since boomers are going to be around for a bit longer in the workplace, how do you get decision-makers to embrace Actionable Data Storytelling? The arrival of generative AI has turned everything on its head. Boomers can embrace data and AI, or they can retire.
If you want senior management to care about being data-informed, you must express excitement and enthusiasm. Find out how they earned their bonus, and show how data can help them. Some will be responsive others won’t. Put on your salesperson’s hat and whip up internal marketing to sell the idea of being data-driven.
Don’t wait for your older managers to move on. Run workshops, do a lunch and learn. Show them how it could be rather than how it is. An old saying is that you don’t get people to build boats by telling them how to build a boat, instead, you should tell them of the majesty of the ocean. Use the same approach on senior management.
Conclusion
They are my best twelve points on how to build an analytics team. By nurturing digital analysts now you can reap the rewards later. Whether you’re a manager, a team member, or a team member who wants to be a manager, why not consider introducing some of these hits into your approach?
About Jim Sterne
Jim Sterne sold business computers to companies that had never owned one in the 1980s, consulted and keynoted about online marketing in the 1990s, founded the Marketing Analytics Summit in 2002 and co-founded the Digital Analytics Association in 2004. His twelfth book is Artificial Intelligence for Marketing: Practical Applications. He is now advising companies on analytics strategy planning at Data Driven Leaders Studio as well as teaching the ins and outs of AI and ML to marketers.
- What is a Customer Data Platform (CDP) & How to Use One - 25/04/2024
- What is Google Consent Mode V2 & How to Implement It - 24/04/2024
- How to Implement Consent Mode with Cookiebot CMP - 09/04/2024