Tuesday, February 25, 2014

TerraSpark - great technology looking for a home

I had a nice short respite after Object Reservoir - I called it a sabbatical, and then started the game again with a company named TerraSpark.

This was a phenomenal technology that came from years of research that started with ARCO back in the 1980s.

The company itself was a spin-off of the University of Colorado, which is where the technology ended up for what you might call an incubation phase. 

There were essentially three bits of technology that could be mixed and matched in order to get a much better idea of what was going on underground in almost any circumstance. Well, as long as you had seismic data.

The first let you automatically pick faults (fractures with displacement) in your data. You need to think about it for a second, but this is harder than it sounds. Faults are really the absence of something rather than the presence of something. When a geophysicist is trying to pick "horizons" or a given formation that may represent oil bearing rock, you generally see a well defined reflection in the seismic data. (Think of seismic data as a sonogram. The horizons are equivalent to the baby's head or bones, or whatever you see in there). Faults exist where the horizons are broken. It is hard to come up with computer rules that let you automatically decide where these breaks take place.



The second lets you "wrap" a geobody (think of a big intrusion of salt) from the outside (shrink wrapping) or the inside (like blowing up a balloon). Again, the hard part is building the computer rules to make this happen quickly.



The last, and most amazing technology lets you move the seismic data back in geological time, to let you see what the seismic would have looked like when the formations were being laid down as beaches, river beds, or deltas. This originally would take days (thing of taking a bunch of pieces of paper, rumple and tear them, deform and invert them, and then algorithmically restore them to flat pieces of paper. Wow!)



As with most new technology, a big part of the problem we faced was convincing our clients that this was something they NEEDED, not something they WANTED. Since they were already finding oil, this was a tough call. 

Another big problem was the performance. 

Computers needed to catch up with the software. It did, but by the time it did there was not much "runway" (some call it cash) left to spend on sales and support.

My role in the company was two-fold - see what we needed to do to increase penetration of sales, and to find a home for the company. I have three transactions in two years, and now I needed to capitalize on that experience.

We had to reorganize the company just a bit so that we could last longer with the cash we had. We also moved the financial part of the company down to Houston so that I would have better and faster access to that information.

And then we started a "process" to get the company some attention.

We hired an investment banker to help with the process, and then contacted all the likely acquiring companies to spark some interest. That was started in the summer of 2013, and by the fall of 2013 we had three "Indications of Interest" that we could work with.

It was an interesting process, but in the end we had CGG Jason as the obvious best fit for our technology.

CGG was to take almost all of the employees (15 of 19), the assets, the leases, and the furniture.

Those negotiations took another couple of months, but by January we were ready to close.

On January 15, 2014, we had our closing conference call. This makes 4 transactions in 40 months for me.

My employment was terminated on 2014.01.31

Monday, February 24, 2014

Object Reservoir. Two sales in one

I stayed with TIBCO for a respectable amount of time so that the technology could be transitioned to the new organization. I have not worked for a big organization for quite some time, so I figured it would be best to move to another small company.

The company was Object Reservoir, a software company that used Finite Element Analysis [FEA] to perform oil and gas reservoir simulation. (Most software used Finite Difference Analysis, as it is "cheaper" from a processing standpoint) It was founded by an old friend of mine several years ago and had sort of a tough time throughout the years.

My task was to see if we could pull the company out of a skid, or lacking that, get what we could get for the technology.

The technology was really quite spectacular. I would like to say that we moved on from the FEA an analytic solution to what was essentially a Big Data problem. The company had come up with a science based analytic and heuristic solution to the question everyone was asking at the time - how do you best produce unconventional gas wells?

We were Big Data before Big Data was cool.

But then gas prices collapsed, and we collapsed with it.

The company had built up a very nice consulting business based on our technology. And besides that, we were able to convince several oil companies to share their gas production data (very proprietary) so that we could perform these analytics across reservoirs so that we would not be influences by local anomalies .

These engineers were good at what they did, but more importantly, they were reservoir and petroleum engineers. We were able to get a small bidding war for that part of the company - and we ended up selling them (as a unit) to Weatherford.

That gave us some time to focus more on the software side of the business.

One problem we faced, however, is that we still hadn't solved the problem of either wet gas or oil. The physics are hard, and though in conventional reservoirs solutions have been found, there has simply not been enough data (big or otherwise) to determine the accuracy of the algorithms.

But we dedicated our scarce resources on solving these problems - but it all sort of caved in at the end.

So we started another "process" to get the software a good home.

It was a short sales cycle, and we were able to work with Halliburton to put the software, and the people, at the Landmark Graphics division of that company.

That was two years that was a lot of work, but ultimately a disappointing end.