Taking a Capital Efficient Company Public and Beyond: Medidata CEO Tarek Sherif (Part 3)

Sramana Mitra: Let’s go back to when you decided to get together and start this company. What was the idea? What did they want to do?

It sounds like they brought the technical and domain knowledge about the CRO industry. What did they have? What was in their mind that they wanted to do?

Tarek Sherif: What was in their minds was that the process of running clinical trials was very complex and a very manual process, in particular, how you gathered and managed the data.

Back then if you wanted to run a clinical trial, you would send these big notebooks filled with the trial protocol to doctors. The doctor, every time a patient came in, filled out the sheet of paper, oftentimes in duplicate, that said, “The patient’s blood pressure was X. I administered a certain amount of drug.”

Those sheets of paper were faxed or mailed to pharmas where people would sit in a room and double data entry all this information into a database. Then there would be reconciliations.

If there were errors, you would send a letter to the doctor. You couldn’t call them and say, “There was an error in your data. Could you please fix that error?” You get a letter back. The doctor will have to go into their file cabinet and find the patient’s chart a year later and look up the data and say, “It wasn’t right or it wasn’t transcribed right.”

It’s a very inefficient process to get data to the point where it could be in a form where it could be handed by your biostatisticians to determine whether your drug actually worked or not. That could take upwards of two years.

So the idea was very simple. You had this brand new technology called the internet. You had web browsers. Why couldn’t you create an electronic version of the form and embed logic into it? If there was an error when you’re inputting the data, you got an error message back saying, “You enrolled a patient who’s too young to actually be in the clinical trial.”

Whatever the error may have been, you therefore get to the point where you could deliver a database of high quality data much faster to the biostats group. It cuts off anywhere from six to nine months to two years of wait time from when your last patient left the trial for the last time.

That was a pretty simple idea but nothing in clinical trials is simple. It’s always very complex. Obviously, people’s lives are at stake, so it is a regulated industry. But the business model and the opportunities seem to make a lot of sense.

The approach we took that was perhaps different was, we decided that you didn’t need to write bespoke software every time you ran a clinical trial. That’s what was happening. People were programming for each new trial. We said, “They’re at a meta level.” Every clinical trial is the same. It’s just that there are variables throughout that you need to change.

The other thing we said is, “It’s very cumbersome being in a doctor’s office with multiple computers.” Back then you had multiple computers. So every time you started a trial, there was another computer. So you would sometimes have eight computers in a room because they were running a trial.

We said, “That’s silly.” If Amazon exists and you can order a book on Amazon, you should have UX or UI that’s as simple to use as Amazon and is low bandwidth. Back then, people were putting DSL lines, believe it or not, to run a clinical trial. You can imagine what the cost and complexity of that is and why doctors perhaps wouldn’t be quite as enamored with using that.

So that’s what we built. A company that had a very user friendly UX – user friendly to the doctor – and that had a library of reusable pieces that represented the commonality within clinical trials. Now some of that sounds like it was genius on our part. Some of it just happened to be luck on our part in reality.

A subscription business model came out of all the same reasons that I’m about to describe, which is we started Medidata in late 1999. We did a friends and family round and raised a little bit of money. We went and built our first product. We were starting to get customers here and there.

We always thought we should start small with a small biotech. We were seven people at the time working out of a very small office and wanted to build a name for ourselves. Then the internet bubble crashed. So there was no funding in New York City for tech companies.

In fact, that was a bad word. If you were in the Valley, there was still some hope. In New York, there was no hope of funding. So we had to get very smart about how we built our tech and how we charged our customers because we were so small and tech was a bad word back then that you couldn’t expect a large upfront payment from a customer for your software.

So we started charging on a monthly or quarterly basis and, ultimately, that led to our subscription business. We built a lot of the tools or the philosophy around how we built the software. We couldn’t afford a lot of programmers. We had to make it very easy to build a clinical trial in our software so that pretty much any one of the people in the office could do it.

That led to a very different approach, as I said, from other companies who were programming each clinical trial and having massive quality issues in that process. So it took them a very long time to do each setup and each clinical trial.

We were much speedier out of necessity. Then the last thing is, since we couldn’t afford to install software behind people’s firewalls and contracted with DSL lines in, we built a central server with our software on it. Therefore the cloud was born.

Sramana Mitra: You were doing multi-tenant back in 1999 before cloud computing became popular.

Tarek Sherif: In 2000, yes. We had the business model around it. It was all out of necessity because we had no money. Less than a million dollars does not go very far in building a company, especially when you’re just getting started.

One of the little known facts about us is, we were always so capital efficient that by the time we went public in 2009, we’d only raised about $13 million and we still had $10 million on our balance sheet. So we built what was basically $140 million revenue business of which over $100 million was recurring revenue.

This segment is part 3 in the series : Taking a Capital Efficient Company Public and Beyond: Medidata CEO Tarek Sherif
1 2 3 4 5 6 7