How to find that first big customer

This is part of an ongoing startup advice series where I answer (anonymized!) questions from readers, like a written version of Smart Bear Live. To get your question answered, email me at asmartbear -at- shortmail -dot- com.

Freshman Salesman writes:

I’ve read somewhere in your blog about how you had a very large organisation as the first customer for your software. I’m putting myself in the same boat now with the solution I’m developing so could you tell me:

1. How did you reach out to your first customer?
2. How you convinced them to pay up when code review at that stage was nonexistent?

I think you meant this?

It’s rarely true that your first customer will be big. It wasn’t true for me either.

You won’t like my direct answers:

  1. AdWords. Here’s the full AdWords story. But this was 2002 when AdWords was affordable and I had no competitors, so you can’t repeat that — it doesn’t matter how I did it.
  2. I didn’t. I didn’t reach out, I didn’t know they reached the website, I didn’t know they were trialing, I didn’t call them to see how it was going, I didn’t convince them it was worthwhile. I had a $39.95 product that barely worked, no sales organization (except tech support, which often works better anyway), and I just hoped they’d run a credit card.

You have a more fundamental issue which, if remedied, will lead to the answers you seek.

To see the problem, let me start with possible solutions. Either of the following scenarios makes sense for a new startup going after a large-scale enterprise market:

  1. You’ve identified a deep pain at a large company. You land a contract to build them the solution. (Maybe you worked there or a co-founder or investor has some juice). The contract says you retain the IP and are allowed to sell a product like this to other companies. (Often they’ll agree if it means a steep hourly discount or cheaper maintenance costs.) You build the product, then you have your banner customer and go get others. In the worst cast, you can’t sell to others, but at least you made money.
  2. You think you have a great idea. You hit your network and the road, calling twenty people at large companies, all of whom are in your potential market. You can’t get to all the decision-markers, but you can find people who are influencers (here’s how) and who can accurately say whether your product would be desirable, and at what price. After interviewing everyone, you tailor your product accordingly, and when you’re ready for beta you have twenty people literally waiting to use it. Some will buy.

Your problem is that you don’t fall into either of these categories because you haven’t found any customers at all.

If you can’t find anyone to buy it, if you can’t even get those meetings, why do you think things will be different once you’ve built your product? And why do you think they’ll then buy the product?

Actually, you already know they will not buy, because you’re asking me how to “convince them to pay up.” If you had something genuinely valuable, that twenty people agreed they needed, you wouldn’t need to convince anyone.

Don’t tell me “they need to see a working product before they’ll have a chat” or any other typical, lame excuse engineers (including me!) make for not putting customer development before writing code.

You’re just stalling. Get your butt in the field, and you won’t have to ask those two questions.

Add your advice to the discussion section!

  • Anonymous

    Very good advice.

    As a current denizen of a VERY LARGE COMPANY it might be good to tell the guy that the purchasing cycle is very very long and then the payment cycle, as an approved vendor, might be 90 days for the first check.  Oh, and approved vendor might take 45 days….


    PS – I am always willing to help small companies sell into the large one where I work. I’m easy to find.

  • Mark Turansky

    Jason, you’re right that any new software idea should be validated by actual potential customers.  I didn’t find 10 who would buy (if I had the product), but I did talk to 3.  I built a prototype and showed it off to one of them who bought into the idea.  That seed money went towards development and that customer helped me through the beta period.

    Now I’m closing in on the second of those 3 and the third has indicated later in the year would be the right time.  We’ll see.  Only signed contracts count.

    But getting the first one is pretty important, and being able to relate to the first one the problems encountered by the other two (or ten) helps the first one jump.

  • I’ve got an addendum to that question, which could be deferred for deeper analysis if necessary.

    It’s relatively easy to experiment with lead generation strategies in consumer markets, where the inter-user relationships are largely self-similar and you can quickly run lots of small tests.  Unfortunately,  conventional enterprise lead generation ideas often require a much longer lead time (trade shows, getting mentioned in industry magazines, etc),  and word of mouth results will come in lumps rather than smooth curves.

    Is there an elegant way to test whether the flywheel will keep accelerating, or do you have to fall back on the assumption that  at least one of the conventionally suggested marketing channels will catch on,  once you’ve got product/market fit?

    • Whenever the sales cycle is long — which almost always is with enterprise — it’s hard to get end-to-end feedback, and therefore hard to iterate.

      Unless… you figure out a *proxy* for the true goal.

      For example, at Smart Bear I looked at historical sales data and I could tell that a lead with a valid phone number was X times more likely to be associated with a later sale, or an email address from a company was X times more likely.

      Or, in our case we had the red book ( which also turned out to be a tremendously accurate measure of interest.

      I was able to turn it into an average — every book sign-up was worth $X in revenue, every lead of this type was worth $Y.  Obviously those averages were extremely rough, and are completely invalid if you tried to apply to a specific lead.  But as an aggregate measure of one marketing campaign against another, it was very accurate.

      Since things like providing informative material and lead-gen is immediate and close to the point of marketing, it’s easy to at least track it.  

      So, I recommend you start with a VERY ROUGH idea of that value per “thingy,” then track those thingies, then over time replace your crappy estimate with more and more accurate estimates.

      It’s still not as tight as you’d like, but it’s a lot better than nothing. 

      •  You can always tell when people are paying attention to this kind of stuff. Folks – this comment is worth 10x the post. Knowing these things is just plain critical to growth.

  • Very good article!

  • wfjackson3

    My view on an early customer comes from the 4 Steps to the Epiphany. My ideal first customer:

    recognizes they have a problem
    has looked for solutions
    has attempted to build their own solution
    is dissatisfied with their solution
    has money to spend

    If they meet all of those conditions, they are going to want what you have.

  • Diyanaa Mercy

    Nice article! I have some problem.

     But i realize my  problem now as well as looked for solutions…!

    Thank u for your advice…