“Are you gonna hire a bunch of useless salespeople like they have at Oracle?”

This was the first of many memorable interactions I had with Eliot Horowitz. Eliot was the founder and CTO of MongoDB and in late 2010, I was interviewing to come aboard as president. Product-led growth was far from the common buzzword it is today, but the founding team at MongoDB had built a product that developers loved – the same developer love that would directly power much of the company’s rapid growth.

My topic today isn’t product-led growth, but the relationship between a founder, such as Eliot, and a hired CEO (or president, or COO–in my case I started as president and later became a CEO) and the key factors necessary for that relationship to succeed. Focusing on that dynamic is even more critical in today’s more volatile, fast-changing technology markets.

On the surface, Eliot’s question to me years ago was about business models and sales hiring. But it went much deeper. Our discussion was a live experiment on how we would work together, getting to the heart of a start-up’s decisive partnership between a CEO and a founder. The territory we covered that day included:

• Was I open to unorthodox thinking?
• Could I justify my plans on first principles?
• Was I willing to engage with a young technical founder on business issues?
• Did discovering that the founders wanted to challenge the established way of doing things make me excited to join – or want to run for the hills?
• Could I make a business decision contrary to the founder’s views and have us both feel good about the process?

All are valid questions and examples of potential tension points between a technical founder and a new leader brought in from the outside. And how a founder and a CEO work through them could help determine the ultimate success of a company.

Beyond product-market fit

Lots can go wrong with a startup, and to succeed, two things have to go right: The product must fit the market well– almost always the domain of the founder(s). The company has to execute successfully–sometimes the domain of a hired CEO.

In almost every case, the initial product and market vision come from founders. They started the company because they had an insight that something could be done better, and an idea of how to do it better. When that idea resonates with a broad audience, you have the kernel of product and market fit. Without that, there is no company.

But that initial product-market fit isn’t nearly enough. A company needs funding, it needs a team, and, ultimately, it needs to execute on engineering, sales, customer success, and marketing. In some cases, a founder is interested in and has shown initial aptitude for leading all these areas. In other instances, they don’t. In the latter case, they need a partner to lead the company’s operations.

The four years I spent at MongoDB–first as president, then as CEO—were a great experience. The company grew explosively and changed the market for databases and how developers built web applications. And perhaps more importantly, we laid some of the foundations for what later became a hugely successful cloud business that transformed how enterprises delivered and consumed infrastructure software.

We wouldn’t have been able to do that without a strong partnership between the founders and me, particularly with Eliot (founder and CTO) and Dwight (founder and initial CEO, who eventually became chairman and handed me the reins of CEO). Decisions didn’t neatly divide into categories of product for them and operational for me. No standard playbook for product-led growth, or for the cloud-based delivery and monetization of open-source software, existed. We built the playbook as we built the company, making decisions about what the initial cloud offerings would look like and which features would be free vs. paid.

Because Eliot and Dwight anticipated that the business model would change along the way, they knew they needed a partner willing and able to combine their first-principles thinking with operational experience to chart a course with them through unknown territory. Eliot’s initial question to me back in 2010 was necessary not because he was looking for some brilliant answer about boosting sales, but because he needed a partner willing to explore how to build the business differently and better.

In the best case, a strong partnership can pioneer new models and build a lasting and impactful company. In too many instances, however, a weak partnership gets in the way of a company living up to its potential in the market.

Here is some specific advice for both founders and outside CEOs, based on my experience as a longtime technology executive and a hired CEO, to get this dynamic right. While the advice is simple to state, it requires self-awareness, hard work, and discipline to execute well.

For Founders:

Be clear about whether you want to run the company or not.

The most important decision that the board makes is who runs the company. You are on the board and need to be front and center in that conversation–and that conversation involves both talking and listening. If you think you want to run the company and the board doesn’t agree, make every effort to understand why they think a change is necessary. If you think another structure could work better, try to understand why the board doesn’t agree.

Less commonly, some founders don’t want to continue as CEO, but the board thinks they should. Again, seek first to understand the views of your advisors.

There was a time–roughly a decade ago–when it seemed in fashion to replace founders as CEO. The industry seemed excited about Godfrey Sullivan joining Splunk* and Frank Slootman joining ServiceNow. If we go back further, people thought that having John Sculley take over Apple was a good idea. Nowadays I believe the investing community–and most boards–recognize the value of founder leadership. The successes of Microsoft, Oracle, SAP, Salesforce, Apple, Google, and Facebook likely helped this viewpoint settle.

There is a lot to say on this topic, and for now, I’ll implore founders not to go into that crucial partnership with hidden resentments, jealousy, or frustration.

Pick a partner you respect.

There are two common pitfalls to avoid here. The first is hiring a COO or president who is “good enough”, but someone you don’t really see as a true partner. The test is when you disagree–if your natural instinct is to try to figure out how to convince this person that you are right, you don’t have the right partner. You need someone who, when you disagree, genuinely wonders whether there is some information one of you is missing, some other answer that both of you will like, or failing that you aren’t sure which of you is wrong.

The other pitfall is hiring a CEO whose resume looks impressive, but with whom you haven’t actually gotten “down and dirty” in discussions, and whose thinking and logic you don’t respect.

Commit to the partnership.

Once you’ve chosen the partner, you’re in it together. When they are wrong about something and you were right, that’s not a “win”. You succeed or fail together.

Partnerships take effort. Go out of your way to notice–and comment on–meaningful things your partner has done well. Ask what you can do to help them.

Advice for CEOs (or top operational executives partnering with a founder):

Pick a partner you respect.

Same idea as above. Is the founder brilliant? Does he or she have a deep–even obsessive–understanding of the market and real insights into what users need? Is having this person at the company a clear competitive asset?

If not, don’t take the job. A founder that is just pretty good is unlikely to succeed. Life is short; don’t waste it working with people who aren’t great on ventures that aren’t likely to succeed.

Don’t put the founder in a box.

Remember that the most successful companies are led by founders. But for whatever reason, maybe you are the CEO. Think of that as an obstacle to overcome on the path to your company achieving spectacular success and try to maximize the contribution of your founder(s) as a way to have the type of success that founder-led companies routinely enjoy. Keep in mind that someday the founder might be CEO (in some cases for the second time), as with Steve Jobs or Larry Page. Your job is to get the most out of them, and in some cases to prepare them to lead the company eventually.

If you ever doubt that’s your role, see above. You’re working with someone brilliant, whose presence at the company is a key competitive asset.

It takes two, and it’s on you.

Partnerships take effort, and both parties have to be committed to making it work. But you need to take the lead in making it work for two reasons:

You’re running the company. Part of your job description is keeping the team aligned. You’re the generalist—your partner is the genius. Yes, your genius partner needs to try to be a team player, but you need to show the way. You wouldn’t have the job if you weren’t good at connecting with people, so put those skills to work! Ultimately, the CEO has a few simple jobs:

1. Not running out of cash
2. Building a strong team
3. Keeping the team pointed in the same direction.

Fundraising and hiring are periodic activities. Keeping the team pointed in the same direction is an everyday focus.

A company founder and CEO need to understand and value each other’s complementary skills. Founders need to have a critical awareness that the skills others bring to the table are valuable. Founders often possess deep market understanding, superb technical skills, or both. CEOs are often generalists or experts on selling and marketing.

It’s easy to think that the thing you don’t understand is easy–or at least not to appreciate the difference between someone pretty good at it and someone great at it. In the end, trust and respect will pay as many dividends down the road as product excellence.

In the next installment of this series, I’ll focus on other critical relationships and roles – such as hiring your first sales leader, and how CEOs and founders can frame their company’s story at various stages – for both founders and CEOs.

This article originally appeared in TechCrunch.