Background decoration Background decoration

Dear Corporate, here is how NOT to work with startups – part 1

small boat navigating towards a big one

In this article I try to convey some of the learnings of the last 7 years at the head of a start-up trying to bring a deeply integrated SaaS to very large1 companies. We have been asked many times to talk about the startup perspective on the collaboration between startups and corporates, so here it is. 

First, for a little bit of context: Our company GenLots recommends large manufacturing companies when to order how much external production materials using proprietary AI. For this we use data from the ERP such as SAP. 

For us, naively, this looked like a no-brainer when we started. 1) the impact is very big ($ millions of potential cost savings per year), 2) integration is fast (within weeks), 3) no current processes have to be changed. 

The view from the corporate perspective, unfortunately, has been very far from a no-brainer. We had many objections such as 1) “It is risky to work with start-ups, especially with one which integrates deeply with our ERP, the cradle of our most sensitive data”, 2) “This is not on my 2 year IT roadmap, how am I going to explain this to my boss and do the project in parallel with 5 others” and  3) “I first have to have good master data and do X, Y and Z before optimizing orders” . 

We have had a good run of over 40 pilots,  large and small, some very large global implementations and many, many disappointments. But, I am happy to report, disappointments have become rarer. So what works? What are the most frequent errors that we have seen? How can corporations and startups set themselves up for success?2 There is a lot of good material for startups3 out there on how to improve collaborations with corporates. But what can corporations do? This is the first article in a 2-part series. It is what I tried to summarize in a (hopefully humorous) step-by-step guide on how not to do it. Followed by part 2, a more serious list of things that we have seen working. 

A step-by-step anti-guide

Let’s make a big assumption. You are working at a large corporation and you are responsible for bringing external innovation to your company. There are startups out there that can bring your company enormous value and you are willing to find them. Here are step-by-step instructions on how NOT to do it. If you want something more upbeat, read the step-by-step instructions on how to DO IT afterwards. 

Step 1: Establish first contact – Innovation by osmosis

An interesting approach guaranteeing lowest concrete outcome for the effort.

  1. Invest in an accelerator4
  2. Co-select with them promising start-ups. Promising is defined as “sexy” according to  the latest fad in the enterprise or the wider world. So if you have new sustainability goals, select many companies which have green tech in the name, even if it has nothing to do with your core business. 
  3. Organize a corporate road trip to a pitching session where the hope is to somehow get “innovation by osmosis” 
  4. Vaguely promise the start-up you will “look into it” and “get back to them”
  5. Congratulate yourself. Innovation has a big priority at your company

If you now think that this is a bit extreme – it is. But it also comes quite close to how many companies organize their innovation days. There are many people with very good intentions – but as a startup it can often feel like being in a Zoo at such events. People look from the outside at those “innovative startupper nerds” without any real intention of interaction. 

Step 2: Get to the first real meeting(s) – People Ping Pong

Let’s assume your magical unicorn of a future value bringing start-up has survived the previous phase and is very persistent with email follow-ups, calls etc. It convinces you to take action. You are now responsible for bringing the startup in. 

  1. Think about all the people they should probably talk to.
  2. Forget to add IT to the list
  3. Organize around 6 meetings in total, but take 1-3 months5 in between because of lacking availability
  4. One year has passed: You can restart from the beginning as important positions have been replaced with new people

My guess is that the people ping-pong until momentum is lost is one of the most frequent stops to corporate-startup collaboration. 

Step 3: Start a pilot – A sure way to lose momentum

We all know that POCs (Proof of Concepts) are important. Start small and scale up etc. You have read the books by Eric Ries and are empowered. So what you decide next, is to do a pilot.

  1. Plan to do it “on the side”. Colleagues are busy, surely they can spare some time for this small project.6
  2. The goal of the pilot is… to see if the start-up actually delivers what it promises.7 Do they have the potential to deliver value and does their technology actually work?
  3. Set-up a timeline which lasts around 9 months – go over by 3 months due to vacation and then report to important stakeholders

An unfocused pilot which is done to “validate” vaguely a startup can be extremely detrimental to the momentum of the project, as it interrupts people in their busy daily work to “just look” at a startup. What to do instead is described in part 2.

Along the way: See internal objections either as absolute blockers or as a nuisance

During all the process outlined above, if one person raises their hand to say something like “ I’m not sure this fulfills our security requirements” or “I think we can do this internally”, or another which is liked by procurement “this needs to go through a formal RFP process” do one of the two following actions:

  1. Take this statement at face value, don’t investigate and abort the whole process
  2. Or: ignore all internal guidelines because “startups are different” and push through the change

The first approach is a show-stopper for obvious reasons. The second one might work for a short while but will run into problems at some point.

Optimism in part 2

Ok, now that I have thoroughly depressed you, look forward to part 2 where I try to find out the secrets behind large corporations which are extremely efficient in working together with startups.8 Maybe you can make your employer be one of them?

→ Here to part 2





  1. If you as a startup somehow can target mid-size businesses, then you do it! Decision structures are much easier. Unfortunately it is in the nature of our product to be only adaptable to large enterprise
  2. I am talking about the one where a SaaS Enterprise startup tries to bring its product to a large corporate client. There are many other types of collaborations like co-selling and co-innovating. For co-selling we have a successful agreement with SAP, on the other hand we had very disappointing interactions with top consulting firms which in the end just want to have you in their database. For co-innovation we have no experience, so it will not appear in this post.
  3. Example:
  4. This is just an example on how to establish first contacts. I think accelerators are generally a good way to get into contact, as they are experts in already pre-evaluating start-ups for general viability, which is not your expertise. But you have to do it with the right attitude.
  5. Except if you are in Europe: No meetings during the 3 months of summer and between 1st of December and 31st of January: important stakeholders are on vacation. So you will have to cram in the 6 meetings with 1-3 months interval into the 7 remaining months. [End of mini-rant]
  6. Side-projects in companies without dedicated people, plans and time often fail. Why should it be different when collaborating with a start-up?
  7. As much as it pains me to say it (As it doesn’t help startups which search for their first client), normally someone else should have already validated that the startup works in principle. We have a first client. Ask us to talk to this client!
  8. Yes, they exist

GenLots stories