How to recruit developers for your startup

Recruiting developers was a challenge for us at Dinero in May 2014. We needed experienced people, but as a small company we could not simply win on the same terms as large employers.

Recruiting developers for accounting software required experience

I sat outside after parking the car and told another story from working on Dinero. I knew some were tired of hearing about Dinero in my videos. But when a project took so much energy, it was also where I found inspiration. I wanted to advertise it, but also share something other startups could use.

It was the first project where I really needed senior developers. In accounting software, we could not simply experiment with people’s bookkeeping and fix mistakes later. We needed to calculate correctly.

I compared it with Amino and Agera, where some mistakes could be corrected afterwards without the same consequences. Perhaps something was displayed incorrectly or a service was down for half an hour. That was not good either, of course, but calculating a customer’s VAT incorrectly was different. So we needed experienced people who could take responsibility for that.

We could not win on the big company’s terms

My picture was that many senior developers already worked in big corporations with high salaries, pensions, holidays and various benefits. I mentioned Danske Bank and talked about massages, soft drinks and seven weeks’ holiday.

I also threw salaries of 50,000 kroner and upwards into the comparison. Those were my examples from 2014, not a survey of actual employment terms. Later in the video, I also stressed I had no particular knowledge of Danske Bank’s technology.

The point was that a small startup could not simply outbid a big employer on all the same things. I could not offer a large gym, the same holiday entitlement or a developer salary of 65,000 kroner a month.

We had created a jobs page with a team photo and a description of our benefits. In my assessment, the pay was not bad, and we also had cola, a massage scheme and free lunch. But it did not bring many applications. Some came, but not what we hoped for.

An idea on Twitter gave me a different angle

I followed some Twitter conversations and came across the CEO of a Silicon Valley startup. He was asked how his business could compete with Google for developers when the big players offered both high pay and all sorts of benefits.

I recounted his answer: When he met a good developer, he asked what they had imagined when becoming a developer. Was the dream to sit in a big corporation keeping something old running, or to write new code?

New code. That was the key phrase for me. It became the angle I used, and I found more interesting applications coming in. I was describing our experience, not a controlled experiment or a guarantee others would get exactly the same result.

My caricature of a big corporation concerned technical debt

I described, very polemically, a developer keeping old platforms running. An older Windows version, a backup system not yet worn out, old servers and an architecture where the new needed to fit the existing.

In that hypothetical environment, a good idea might meet: We tried that. It does not fit the old system. Customers are not ready. You used your skills to keep the machine running and only occasionally got to build something new.

I exaggerated the picture, talking about sitting with a pot belly and Danish pastries, simply waiting for Friday. That was my provocation about professional stagnation, not a description of every developer or big workplace.

My concern was that you could lose your grasp of new technology if work constantly concerned the old. If the company later changed systems, your skills could have become less relevant.

I thought employees had a responsibility to stay sharp. But I also saw the difficulty when the employer constantly required work on old platforms. Then some learning had to happen in free time. That was my view of the challenge then.

We could offer influence over what would be built

I did not understand all Dinero’s technical details myself. But I was excited that we worked with a new Microsoft platform, apps, HTML5, backup and an API-based structure where things could connect.

When we saw exciting new technology, we could investigate and play with it to see whether it was useful. That was the environment I wanted to show better in our recruitment message.

I also mentioned startups such as Autobutler and Agera. My point was that in a new company, you could write new code and help define how the business would work in future. There could be fewer old constraints to consider, what I called technical debt.

I thought that opportunity could excite a developer enough to accept lower pay. I contrasted 65,000 and 40,000 kroner. Those were illustrative amounts in my argument, not documentation of what any particular employee accepted.

In my view, the extra pay also had a price

I also repeated the thought that Google had itself become an old company with much old code. I compared it with the big Danish corporation as part of the same argument.

My sharp wording was that an extra 15,000 or 20,000 kroner a month was not necessarily free. If in return you learned nothing new and lost professional sharpness, there could be another price.

That was my opinion in the video. I did not know the specific work of all the developers I discussed. But I wanted candidates to think about both pay and what they would spend their days doing.

Users’ reactions could give work energy

I used an image of writing a procedure deep inside a big system. If it worked, perhaps nobody noticed. Important work could be invisible precisely because you had prevented something breaking.

At Dinero, I imagined the opposite: You wrote a new feature, it was used and then emails came from happy users. I used perhaps 50 emails as an image of direct response, not a count from a particular launch.

That energy could keep you going. You went home and found new possibilities online, talked to colleagues and looked forward to trying things. That was the enthusiasm I wanted our recruitment to show.

So my encouragement to other startups was to communicate clearly about new code, influence, technology and the chance to make a difference for users. To developers, I encouraged considering where they could stay sharp and feel like building further.

Recruiting developers with a different message

We began with a jobs page describing the benefits of working with us. It did not bring as many relevant applications as we hoped.

A Twitter conversation gave me another angle. Perhaps we needed less space for benefits and more for the work a developer could actually do.

New code and a chance to influence the product

We began talking more about the technology, building something new and having influence. In a small company, you could be close to both decisions and user feedback.

That mattered particularly in accounting software, where the code needed to work properly. We needed people willing to help take responsibility for the product.

It was our experience at the time

I found the changed message brought more interesting applications. That is my description of our process, not a controlled comparison or a guarantee for other companies.

The video’s pay and holiday examples are my assumptions from 2014. Nor did I have specific knowledge of technology at all the big companies I mentioned. The story is about finding what makes the work special and explaining it clearly.

Also read Min vej som leder for more of my experiences with employees and leadership.

Also read about practical tasks and learning in small businesses.

Recruiting developers: The next task was getting the message out

In the follow-up, I said we had received around 15–20 applications at Dinero. About a third looked interesting. The process was still underway, but I hoped we would find the two or three people we wanted.

I was no recruitment expert. I shared what we experienced and my ideas for next steps. First, it had been about explaining to experienced developers what we could offer. Now it was about getting the jobs page to the right people.

The message needed to convey enjoyment of work

I returned to the joy of creating something new. Would you prefer higher pay and work with old code and limited influence, or exciting projects with skilled colleagues and a role shaping your own working day?

I knew a startup involved maintenance, too. But I thought there was a difference between joining an old organisation with many existing systems and a new company with people passionate about building something.

Of course, a startup also carried risk. I mentioned the possibility of the business failing and needing another job. My assessment then was that a skilled developer had good chances of finding one. That was my 2014 assessment, not a guarantee for the individual.

A team photo could show the energy

I thought Facebook was interesting because developers were there, too. I thought you could reach a fairly precise audience at a relatively low price, but repeatedly stressed I was no Facebook marketing expert.

My suggestion was to use a company page or personal page, depending on what made sense. I would probably start with whichever had the most likes.

Then I would have a good team photo taken showing the company’s spirit. On Dinero’s jobs page, we had a picture of developers around a Scrum meeting. They stood discussing the work, and I thought you could feel the energy.

The picture needed a headline catching a developer’s attention. My example was: Are you tired of writing old code? It should touch what could make the job itself interesting.

My thoughts about Facebook were from 2014

I discussed Facebook’s Power Editor at the time and putting a status update into the advertising tool to target particular audiences. I was unsure of details, including which browser and installation were needed.

So those were my thoughts on the tool then. I was not describing a tested recipe for every step, and the old description should not be read as guidance for Facebook advertising today.

I also suggested narrowing by gender and age. I used a very high figure for the proportion of male developers and mentioned various age limits from the mid or late twenties to forty or fifty.

That rested on my assumptions about whom you could reach effectively, and about age and interest in new technology. I did not present research supporting them. I also said skilled developers existed both younger and older than the limits I mentioned.

I knew the statements could attract criticism and changed my suggestion along the way. They are reproduced here as historical thoughts in the video, not a current recommendation to assess qualifications or recruit by gender and age.

I would include geography, too. I imagined getting a developer in Jutland to move to Zealand could be difficult. So I would consider where candidates lived when setting up the campaign.

Ask developers where they find professional knowledge

I would also try similar things on Google+ and buy Google AdWords. But which search terms should show the adverts?

Here, I would ask developers what they searched for in their work. Perhaps particular code libraries or other professional topics the target audience used when seeking knowledge.

That could suggest keywords and programming communities or other places relevant people visited. It was a suggestion for learning the audience’s habits, not a result from a completed campaign.

A tip about employees’ networks

I had also received an idea from someone I knew: Ask existing developers to suggest skilled people in their networks.

The specific idea I recounted was to collect names and email addresses of potential candidates without linking suggestions to the employee providing them. A recruitment agency would then handle contact.

It was the connection back to the employee that would be anonymous, not the candidate’s name. The idea was that the employee would not appear to have lured an acquaintance away from another workplace.

I shared the advice because I thought it an interesting way to use the network more broadly. I did not say we had done it, and I did not review the conditions for collecting and using contact information.

I wanted to hear more experiences

Of course there were also job sites, headhunters and other options. I had little experience with them, so I would not pretend otherwise.

I ended by asking others for ideas on places to find developers and other ways to do it. We had a better message and some interesting applications. Now I wanted to learn how to reach more of the right people.

Get the recruitment message to developers

In the follow-up video, I said we had received more applications and were still in the process. I also wanted to discuss how the message could reach relevant people.

I suggested a photo showing the team and energy in the work, with a headline about the development task itself. I considered social media, search adverts and places developers found professional knowledge.

I also recounted an idea for using employees’ networks through a recruitment agency. These were suggestions in 2014, not documented results from every channel.

My statements about gender, age and particular advertising tools are historical assumptions. They are not current guidance on setting up adverts or assessing someone’s qualifications by age and gender.

Historical job reference from 2014: The reference at the time now leads to Dinero’s current jobs page.