Planning for Your Startup: The Data Team's Guide to 2021

I love this time of year! And no, I’m not talking about holiday parties on Zoom. What I’m really excited about is the opportunity to shape all those big ideas into something that can become reality. Isn’t this why we do the startup thing in the first place?

But planning in a startup can feel like an exercise in futility — especially when it comes to data — especially when your data team is small and scrappy.

I know the feeling. I’ve spent several years working with data, startups, and data in startups, and I know the struggle. It’s a constant push-and-pull to balance working in service to other teams vs advancing your own agenda and truly putting the data to work.

I recently discussed this with a friend and project management wiz, Anders Schneiderman. The conversation, the books Your Strategy Needs a Strategy, and The Lean Startup inspired us to collaborate on this post.

Here’s what we came up with: A guide to prevent you from thinking of plans as nice ideas that are quickly thrown out the window, and to start thinking of planning as a way to build balance and strategy into the way data works at your company so you can help your startup thrive and make your corner of the world a little more sane.

What does it mean to create a plan for a startup?

Before we dive in, there’s a question you need to answer: in your startup, what exactly does it mean to create a plan?

If you worked for a big corporation, you could create a plan for an entire year. And assuming that another turning-our-world-upside-down-crisis like the coronavirus doesn’t hit in 2021, you’d have a decent chance of executing it.

In a startup? If your plan survives an entire quarter, you get a gold star.

A year-long plan doesn’t make sense for most startup data teams. But you’re probably going to spend a good chunk of 2021 being bounced around like a pinball unless you have at least some sense of where you want to go. So how should you construct your plan?

The most useful way to approach creating a plan is to think of it as a roadmap:

• Q1: Here’s what I think is reasonable to expect to get done.
• Q2: Here’s what’s probably going to be happening and what I have a shot at getting done.
• Q3 & Q4: Given what I know right now, here’s what would be nice to accomplish. Or to put it another way, here’s a parking lot of tasks I hope I’ll get to work on in the second half of the year. Take stock

Before you start thinking about what you hope to do in 2021, spend some time analyzing what you can learn from 2020. Here are some topics you’ll want to explore:

• What’s currently working well? What needs improvement?
• From the successes you had this year, what lessons can you learn? What mistakes did you make this year that you’d like to avoid making in the future? And are there any weaknesses you discovered as a result of these mistakes that you’d like to try to improve?
• Are there any areas where you (or the rest of the data team) were a blocker to other coworkers or projects? If so, what can you do next year to reduce or eliminate these problems?
• Are there areas where you accumulated the data analyst’s equivalent of technical debt? For example, are there a bunch of SQL scripts you had to frantically bang out when you unexpectedly pivoted that are important enough to day-to-day operations that you need to go back and clean up and document in 2021?
• Are there any areas where a more self-service approach would have really helped the teams you support?

You can use these prompts to play the game, Start, Stop, Continue.

Whatever you do, don’t use taking stock as an excuse to beat up on yourself. In the chaos that is startup life, nobody’s perfect. The point of looking back isn’t to point out all your flaws, it’s to help you improve.

Similarly, don’t use what you didn’t accomplish as an excuse to ignore or downplay your successes. Celebrating your wins isn’t just good for your ego, it helps get you charged up for next year.

Evaluate your resources

Once you’ve identified areas you want to address in 2021, figure out what resources you’re pretty confident you’ll have next year and what you might be able to get if you make a strong enough business case. Need a guide? Here’s one on how to propose a data stack. The same thought process applies to arguing for all sorts of resources.

Your primary resources are money and headcount you directly control. But you should also consider whether there are resources you’ll either need or could elicit from — or share with — other teams.

For example, if you rolled out a new BI tool in 2020, will some of the teams who aren’t using it have time to get trained up in 2021? And would teams that are heavy users of the tool have the time and interest to work with you to deepen their skill in using it?

It’s also worth asking whether there are any projects where you think you can enlist members of your network of allies to either help out or take the lead. What you think of as just another day at the office, they may view as a coveted professional development opportunity.

Finally, now’s a good time to find out what teams are planning for next year that will require dedicated resources from you. Knowing that in advance will ensure that your expectations are aligned…or at least that you’re not blindsided by someone else calling dibs on your time.

Define your projects and set your goals and metrics

For this phase of your planning, what you need to do depends on how your startup handles developing goals, metrics, and project plans. Regardless of which planning method your startup uses, here are a few tips on how to get the most out of this stage in the planning process:

1. For hard-to-measure work, shoot for good-enough metrics

For some of the most important work you do, you might find it difficult to accurately measure your impact. That’s okay, just use a good-enough proxy.

Say you’re going to use training and mentoring to deepen your team’s ability to use your BI tool. Ask yourself, if they’re getting more skilled at using the tool, how would I expect the quantity and types of support (or Jira) tickets I handle to change? For example, would I expect the total number of tickets to drop, or would I expect I’ll get fewer tickets for simple list requests and more tickets for more complex analysis?

Once you’ve nailed down what you expect to happen, create metrics, and get them into dashboards. If you need to, add new categories or tags to your helpdesk software or Jira account so you produce the data you need for the metrics — e.g., a category/tag for simple versus complex requests — far enough in advance of the new project to establish a baseline.

2. Don’t neglect your own needs

Facing a never-ending stream of requests, many analysts guilt themselves into sacrificing work that benefits them. It’s understandable, but it’s a mistake.

Taking the time to invest in your skills, explore new tech, and build views and other infrastructure that reduce the effort it takes to fulfill requests doesn’t just benefit you, it also benefits your company. So be sure your plan carves out time for projects that make your life better and move the business forward.

3. Don’t rack up too much data infrastructure debt

Make sure your plan includes time to clean up SQL, write or rewrite documentation, add new views that better fit where teams’ needs are headed, and so on. How you do this will depend on your startup’s culture and the challenges you currently face.

For example, rather than setting aside January to knock out that work, you might be better off adding helpdesk tickets throughout the year. Or you might want to add time to some high-priority projects to pay down a little infrastructure debt in the project’s early stages. This is also a great place to get help from your allies: you can ask them to put in requests that pay down the debt that impacts their work.

4. Just because you have to write a year-long plan doesn’t mean you have to believe in it

If your startup requires you to create a plan that suggests you know what you’re doing for an entire year but you’re still in the “next quarter is uncharted territory” stage, don’t sweat it.

Managing a startup is a nerve-racking endeavor, and if the people at the top need to pretend they’ll have more control than is likely, that’s okay. There’s nothing wrong with having a business blankie so long as everyone knows that’s what it is.

Write up the plan

Now that you have a rough idea of what you want to put into the plan, what’s the best way to write it up?

There’s no one right way to write up a plan. Your main goals with your writeup are: 1) to make your boss and whoever else needs to sign off on the plan think it’s clear and easy to skim, and 2) to be able to refer back to the plan in a few months and easily see if you’re on track.

You may want to make a strategic decision about where and how to document your plan. For example, while plans typically are written as Google or Word docs, if everything in your startup ends up in Jira, you might want to check with your boss to see if it makes sense to skip writing a document. Instead, you could create one or more projects in Jira, then create a backlog for each and schedule changes from there. (Note: If you go down this road, you might want to explore using milestones and components as a way of structuring it).

The other thing you’ll certainly want to do is get feedback on an early draft. Obviously, you’ll want your boss to weigh in, but you should also get input from key higher-ups, users, and allies. Getting that input early will help you come up with a finished product that satisfies all your stakeholders.

Conclusion

It’s easy to drive yourself a little nuts when creating a plan for next year — especially in the fast-moving world of startups. But there’s no reason it has to be that way. So long as you use the simple, structured approach we’ve discussed here, you should be able to come up with a 2021 plan that will make next year more predictable (and enjoyable!) for yourself and help your startup succeed.