Last week, I had a conversation with the co-founder of a Berlin-based startup about senior engineering leadership.
I am deliberately keeping the company anonymous. The conversation was simply a useful case study. It revealed several mistakes I see startups making when they try to attract technical leaders—and a few mistakes I made myself.
From the beginning, I could feel that I wouldn’t be especially interested in this opportunity. The company had some appealing qualities, but the overall proposition did not convince me. As the conversation continued, however, I found myself diagnosing its hiring problems, questioning the role it was trying to fill, and offering recommendations about how it could position itself to engineers.
What was supposed to be an exploratory conversation gradually became an unpaid consulting session.
That was their mistake, perhaps. But it was also mine.
A technically interesting company can still look boring
At one point, I was asked why the company was finding it difficult to attract talent.
My first thought was simple: the company was presenting itself as a real estate business.
There is nothing inherently uninteresting about real estate technology. In fact, the industry contains enormous technical problems: fragmented data, complicated workflows, physical-world constraints, large capital projects, compliance requirements, scheduling, risk, and coordination across many parties. Add AI to that environment and the problems can become genuinely ambitious.
But that was not the story being told.
An engineer visiting the company’s website would mostly see construction. They would not immediately see the difficult technical problems, the scale of the opportunity, or why working there might be more compelling than joining an AI-native product company.
That matters because companies do not compete for talent only with their direct business competitors. A construction startup may compete for customers with other construction platforms, but it competes for engineers with AI labs, developer-tool companies, well-funded SaaS businesses, and every other employer offering interesting technical work.
Candidates are comparing missions, technologies, colleagues, compensation, flexibility, and future career value. They are not limiting themselves to companies in one industry.
My recommendation was to build a stronger narrative around the underlying technology and the problems being solved. Productize the story. Make the complexity visible. Explain what is technically difficult, what becomes possible if the company succeeds, and why an excellent engineer should care.
This does not mean pretending to be something you are not. It means not underselling what you have actually built.
If the interesting part of the company only becomes clear after a one-hour conversation with a founder, the company has already lost many candidates. Most of them will never book that conversation.
Your employer story starts before the interview. The website is part of the hiring process.
Five days in the office is not a cultural preference. It is an employment cost.
The second issue was the requirement to work from the office five days a week.
I understand why founders want this. Especially in an early-stage company, they want speed, focus, spontaneous collaboration, and a strong sense of shared momentum. They may believe that colocating everyone is the simplest way to create those conditions.
But that is only half of the equation.
The employer sees intensity and collaboration. The candidate may see a loss of autonomy, family time, travel, and control over daily life. A high salary does not automatically compensate for that loss. For many experienced engineers, flexible work is now part of the value of their employment—not an occasional perk that can be removed without consequence.
Remote work may continue to evolve, but flexibility is not disappearing.
This does not mean every company must become fully remote. It means that strict office requirements narrow the available talent pool and therefore have a price. A company asking for five days in the office needs to offer a sufficiently strong reason for candidates to accept that trade.
The largest and most prestigious employers can impose unpopular conditions and still receive thousands of applications. A lesser-known startup does not have the same leverage. It cannot copy the demands of a top-tier employer without also offering comparable brand value, career upside, learning opportunities, compensation, or mission.
If attracting exceptional people is already difficult, rigidity is not a neutral choice. It makes the problem harder.
The mythical Head of Engineering
The company was hiring a Head of Engineering. When I asked what it wanted this person to do, the description sounded familiar.
It wanted a builder: someone obsessed with technology, active in the codebase, capable of moving quickly, and excited to discuss implementation details for hours.
It also wanted an executive: someone who could design an engineering organization, hire and develop people, establish processes, own budgets, scale the team from five engineers to twenty or thirty, and take responsibility for architecture, security, infrastructure, DevOps, and perhaps even IT.
In other words, it wanted everything.
This pattern appears frequently in startup hiring. Companies want the speed of a founding engineer, the technical depth of a staff engineer, the management ability of a seasoned director, and the commercial judgment of an executive—all in one person, immediately.
Such people exist, but they are rare. Even when a person has all those capabilities, they still have only so many hours in a week. Skills do not remove capacity constraints.
The important question is not, “What impressive qualities could this person have?” It is, “What is the most important problem this person must solve during the next twelve months?”
If the immediate need is to improve the product, make architectural decisions, and increase delivery speed with a small team, the company may need a strong lead engineer or founding engineer.
If the immediate need is to build an organization, create a hiring system, develop managers, allocate budgets, and prepare the company for scale, then it may need a Head of Engineering.
Titles do not solve ambiguity. Sometimes a senior title is used to make an opportunity appear larger while the actual expectation remains almost entirely hands-on. That mismatch eventually disappoints both sides: the founder believes the leader is not coding enough, while the leader discovers there is neither the scope nor the authority implied by the title.
Before hiring a Head of Engineering, founders should be able to answer three questions clearly:
What problem must this person solve now?
What will they stop doing as the organization grows?
What decisions will they genuinely own?
If those answers are unclear, the role is unclear.
Salary transparency does not replace compensation discipline
Another part of the discussion concerned retention. The company had public salary bands, which I respect. Transparency gives employees useful information and can prevent arbitrary compensation decisions.
But transparency also requires discipline.
If an employee says they have received an external offer far above the top of their internal band, the company has a decision to make. It can determine that its market data is wrong and adjust the band. It can make a documented exception because the employee’s role or impact is exceptional. Or it can decide that matching the offer would damage internal equity and allow the person to leave.
What it should not do is improvise compensation policy under pressure.
Reactive counteroffers may retain an individual temporarily, but they can also teach the organization that the fastest route to a raise is to produce an external offer. The issue is not whether the offer is genuine. The issue is whether the company has a compensation philosophy strong enough to make a consistent decision.
Public bands are a good start. They are not a substitute for judgment.
When a discussion becomes free consulting
By the end of the conversation, I had offered detailed opinions on the company’s employer narrative, hiring strategy, office policy, organizational design, and compensation decisions.
I had started working.
This is a boundary I am still learning to manage as I develop my consulting work. I want to demonstrate that I can see problems clearly. I want to build a reputation and create a portfolio of useful thinking. The temptation is therefore to keep giving: one more diagnosis, one more framework, one more concrete recommendation.
But generosity without boundaries easily becomes free labor.
An exploratory conversation should establish whether there is a problem worth solving together. It should not solve the problem in full. A consultant can show the quality of their thinking without transferring the entire value of that thinking before there is any commitment.
Afterward, there was no natural next step. A later comment referred to a technical specialization that had barely featured in our discussion, which felt disconnected from the conversation we had actually had. That was frustrating—not because an informal exchange must produce an engagement, but because a straightforward acknowledgment that there was no reason to keep talking would have been more useful.
Still, I cannot place all the responsibility on the other side. I knew early that this was not a situation I wanted to become involved in, yet I stayed in the discussion and continued offering value. I could have closed the conversation sooner, kept my observations at a higher level, or said plainly that a deeper diagnosis would need to become a consulting engagement.
That is the lesson I am taking with me: do not confuse demonstrating expertise with giving away the engagement.
Candor is not enough
There was another uncomfortable lesson in the way I behaved.
When I am interested in a relationship, I put effort into how my opinions land. When I am not interested, I can become ruthlessly direct. I remain polite, but I spend less energy translating the truth into something the other person can hear.
It is easy to call that honesty. Sometimes it is simply impatience.
Good consulting is not just the ability to identify what is wrong. The diagnosis must be delivered in a way that preserves enough trust for someone to act on it. A correct observation that only triggers defensiveness is rarely useful.
That does not require flattering founders, hiding disagreement, or pretending to be enthusiastic. It requires treating communication as part of the work. Candor without care can become carelessness.
I do not regret questioning the company’s story, its office policy, or the role it was trying to hire. I do think I could have been more deliberate about when to challenge, how much to give, and whether the conversation was worth continuing at all.
The real competition is the complete employment proposition
Companies often ask why they cannot attract the people they want, as though talent were failing to recognize the opportunity.
Usually, candidates are responding rationally to the proposition in front of them.
If the industry sounds unexciting, make the real technical challenge visible. If the company requires five days in the office, acknowledge the cost and offer a compelling reason. If the role combines four jobs, decide which one actually matters. If compensation is transparent, apply the policy consistently. And if you ask a senior person for strategic thinking, recognize when the conversation has crossed into consulting.
The candidate experience is itself a product. It has positioning, pricing, constraints, and a user journey. Like any product, it cannot succeed merely because its creators believe it should.
And candidates and consultants have responsibilities too: recognize disinterest early, protect the value of your expertise, and do not use bluntness as a substitute for thoughtful communication.
The company in this story may go on to become very successful. Early-stage companies are unfinished by definition, and so are many of their leaders. But potential is not the same as readiness—and neither companies nor candidates benefit from pretending otherwise.