Starting a company in the ad ecosystem

(Originally published on ClickZ, July 2007) by Eric Picard

As a serial entrepreneur who’s started a few companies in the ad technology space, I get a lot of requests for advice about starting companies. This is also a byproduct of my current job, which involves much work with ad tech acquisitions. Given events of the past few months, the venture and start-up communities are hungrily looking at advertising as a place to invest.

So today, some advice for both investors and entrepreneurs who are looking at advertising and trying to figure out what to do.

Honor the Ecosystem

Most people starting companies believe they’ll succeed by disrupting the ecosystem. They look at a value chain to see where they can cut a major player out of the mix and capture its value.

In advertising, the focus is often put on disintermediating the ad agency. It’s been tried many times, and it never works. People who don’t work in advertising erroneously think an agency is a creative production shop. Agencies play a huge, valuable, and constantly evolving role in the advertising ecosystem. They also have immense power and are not to be trifled with. They’re not just creative shops. They provide a wide range of services, including strategy, creative, media planning, media buying, marketing analytics, and many others.

Rather than try to disintermediate agencies or other players, look at market inefficiencies in the advertising ecosystem. Where can you provide value? How can you make things easier for companies in this space? Where can you provide transparency for things that are opaque? Those are the ways technology companies have succeeded in the space.

The Entrepreneur’s Formula

Back in the day, a formula was widely used to figure out how to make millions as an entrepreneur. It went something like this:

  1. Raise a few hundred thousand dollars from angels, friends, and family.
  2. Build a team, create some software, get a good proof of concept, maybe get a beta and some strong client relationships, if possible.
  3. Raise a few million dollars from a reputable venture capitalists (VCs), giving away 20 percent of the company (Series A funding).
  4. Hire more people and some experienced executives. Build out a strong version one product and get some customers signed up.
  5. Raise $8-10 million (Series B funding) from several VCs, giving away 20 percent of the company.
  6. Hire a new CEO, get the company profitable, hire more people. Get to about 50-75 people.
  7. Raise another $8-10 million or more (Series C funding) from the same VCs and a few strategic investors, giving away 20 percent of the company.
  8. Maximize revenues to justify a $60-100 million price tag.
  9. Sell the company to Google, Microsoft, AOL, or Yahoo.

The problem with this formula is most of the time, it doesn’t work well for entrepreneurs. It’s very hard to get acquired for $100 million. And it’s hard to meet the formula’s requirements.

It also pushes the founders’ ownership percentage relatively low. And God forbid you don’t sell the company in the first three to five years, because the VCs will typically have dividend payments that grow over time, eating into company ownership. Investors are always paid before anyone else. The amount the company must be sold for so founders and employees see a return is quite high by that point.

Say you’re the founder of a company that followed the above formula. By the time you get through series C, you personally own 5 percent of the company, and you’ve taken $25 million in funding across three rounds. Here’s how it works:

  1. The company’s sold for $60 million.
  2. Investors have a (typical) 1.5 times conversion on their preferred stock: $25 million x 1.5 =$37.5 million. That leaves $22.5 million.
  3. Their stock converts over to common stock.
  4. Investors participate again (typical) as common shareholders.
  5. As the founder, you get 5 percent of $22.5 million, or $1.12 million.

Technically, you’re now a millionaire. But you must pay capital gains on this money at a pretty high rate. So now, you’re not a millionaire.

Think Smaller

Let’s think about this with a new formula:

  1. Start a company with a $200,000 investment from angels, friends, and family.
  2. Build a technology that adds specific value to the ecosystem, is relatively simple, and takes engineering talent and resources to build. The technology should have extremely open APIs (define) and be very defensible to acquire (i.e., doesn’t contain lots of other people’s technology).
  3. Raise a few million dollars from a reputable VC, giving up 20-40 percent of the company.
  4. Hire more developers and one or two killer business development people with strong industry relationships. Get the company to 20 people, very engineering heavy, and complete version 1.0 of the product.
  5. Get some customers to adopt the product, including a big company.
  6. Sell the company for $10-40 million.

What happens in this case?

  1. The company takes $3.5 million in investments and is sold for $20 million.
  2. Investors take $5.25 million off the top, leaving $14.75 million.
  3. Founder owns 30 percent of the company at this early stage (assuming two other cofounders and employee stock options) and cashes out with $4.42 million.

Right about now, you’re thinking, “Wait a second, Eric. Are you saying it’s in the entrepreneur’s best interest to sell his company early for less money? That seems wrong. Shouldn’t you try to build a big company with a complex product set that solves big problems?”

No. The new world we live in is very different. You’re better off building a small company that solves a small to medium-sized problem that’s very technically complicated. Before joining a big company, I erroneously assumed these guys had so many resources that nailing small technology problems was a no-brainer. I’ve found, however, big engineering teams are focused on solving big engineering problems. Big companies suffer from the same problems small companies do: they never have enough resources to do everything they need. The difference is the big guys have money to acquire companies to speed their time to market.

Most startups try to solve the whole problem. They build really fast with a small engineering team and a large marketing, sales, and operations team. The engineers hardcode everything and don’t keep the code open, don’t document the code, and insert all sorts of crazy open-source widgets into their technology. Often, the engineering is outsourced to another country, and founders give very little thought to ensuring security on the other end of the pipeline.

When a big company comes along and does due diligence on the technology, it finds security holes, spaghetti code (define), lots of technical problems that have to be mitigated prior to the acquisition, and lots more that have to be solved later. And if the start-up created an entire operating platform, there’s likely a ton of redundancy between the acquirer’s and the startup’s platforms.

And the outsourced development that sounded like a great idea? Let’s just say if you haven’t visited the team building your product and have no idea what type of security and source code management they’ve used, you might have some trouble. If you’re trying to sell your technology to a big company, it probably isn’t great that every developer in Eastern Europe or Asia has access to bits of your source code. It’s better to build an engineering team locally that’s a strong asset and adds value to the acquisition price. It’s hard to find talented engineers at a big company who have experience in advertising technology. Startups have better luck recruiting engineers, and they can gain valuable years of experience building a version one or two ad technology that makes them more valuable.


The biggest pieces of advice I have, then, are:

  • Find a place in the ad ecosystem where you can add significant value without taking on a powerful adversary.
  • Build a strong technology company that solves a piece of the problem.
  • Expect to exit by selling early for less money — before taking a lot of investor money.
  • Don’t try to boil the ocean. All the redundant code you write that mirrors what an acquirer already has is probably not of great value to it.
  • Focus on where you can create value to the company buying you.
  • Hire great engineers locally.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: