CRN Exclusive: Tintri Co-Founder Harty Talks IPO Delay, Post-IPO Plans, And Why Tintri Is Not Really A Storage Company

Looking Past The IPO

All-flash storage technology developer Tintri was set to hold its IPO Thursday, June 29, but didn't. Instead, it cut its IPO price from the expected $10.50 to $12:50 per share to $7.00 to $8.00 per share, then held its IPO on Friday, June 30. It opened at $7.00 per share and has since traded close to its opening price.

Kieran Harty, Tintri's co-founder and chief technology officer, discussed the Mountain View, Calif.-based company's IPO with CRN, and said the delay was caused in part by investors lumping Tintri into the same bucket as a number of storage vendors that in the past couple of years have stumbled post-IPO instead of thinking of Tintri as a cloud technology provider.

Harty also discussed plans to make Tintri profitable, and offered a peek into product development plans, but cited the fact the company is now public as a reason for not getting into details.

Here's a look at the thinking behind the IT industry's most recent IPO.

Pictured l. to r. are Harty, co-founder; Ken Klein, CEO; and Mark Gritter, architect and co-founder.

Why did Tintri delay its IPO by one day last week?

I can't go into all the details, but I can give you some flavor on this. If you look at the market right now, [it] has characterized Tintri as a storage company. We know this is a time when storage companies have struggled, at least the ones that did IPOs in the last few years. The reality is, we're an enterprise cloud company. And so we're really temporarily held back by the performance of the comps. It's kind of like being a good house in a bad neighborhood. Over time, we will move neighborhoods as we establish what market we're actually in.

So we're addressing a more compelling market. It's a $27 billion market, and we have large enterprise customers, including people like Chevron, United Healthcare, Shire and Comcast. Our customers love us, and we believe over time the gap between the market perception and what our customers think of us will close significantly.

Tintri is indeed thought of as a storage company. Why do you say Tintri is a cloud company, not a storage company?

We started out providing storage as a web service. That's the basis of what we do. [The cloud] is built using web services. AWS is Amazon Web Services. We have a significant set of web services built on top of that including automation, data-protection-as-a-service, and analytics. If you look at what our customers are doing with us, the large customers, including companies like Shire and some other companies that we can't name that are cited in the S-1, are essentially delivering enterprise clouds. And the defining characteristics of the enterprise cloud are essentially self-service and automation. But those are things you couldn't do with a conventional storage product, which is using a very physical model. And so, really, the proof point is with the customers in terms of what customers are doing with our products.

Well, then, couldn't you classify Tintri as a provider of storage technology for cloud services?

A lot of what we provide is software-based in terms of analytics, which are at the VM [virtual machine] level, which goes beyond just storage. So for a long time we've had analytics that allow you to look at post-network and storage latency. We have today predictive analytics that allow you to do planning for storage. Later this year, we will be adding analytics that allow you to do planning for compute as well.

Storage is a wedge for us. It's a very important aspect of what we do. If you look at what happened on the public cloud side, Amazon started out with S3 [Simple Storage Service] as one of their first services. And that was really the foundation for the hundreds of services that were built over time. We believe that if you do not solve the storage problem, you're not going to be able to build an enterprise cloud.

So what actually happened between Thursday, June 29 and Friday, June 30. Nearly everyone was expecting the IPO on June 29, but then it happened on June 30.

We had some logistical issues. Again, I can't go into detail, but there was an additional filing of the S-1 with the SEC on Wednesday, which meant that we couldn't make Thursday. And so we were on track, but we just needed an additional day to actually do that.

In that additional filing, the main difference was the planned IPO price, which had been expected to be in the $10.50- to $12:50-per-share range, but was updated to be between $7.00 and $8.00 per share. Why the drop?

As I said earlier, [we're] compared with other companies. The market bucketed us with the storage companies. We believe strongly that over time we will differentiate ourselves from that group. And it's based on the quality of our customers, and the fact that we have 21 of the Fortune 100 companies. We also have a very significant repurchase rate. Our top 25 customers buy 19 times their initial purchases. And we're 15 percent in terms of software-attach rate.

Still, that was a big drop in the IPO price.

We're not focused on what the IPO price is, quite honestly. We're focused on building a significant business which is based on large customers, and on large customers who repurchase a lot of stuff. And we're also based on having high gross margins. We have the highest gross margin -- 65 percent last year -- of a company of our size in the enterprise infrastructure space. We of course would have liked the price to have been higher. But we're focused on the long term, which is really about delivering products, getting to cash flow break-even, and to profitability. And we have a plan to do that.

According to Tintri's S-1 filing, Tintri's costs are rising with revenue.

Last year, we grew revenue at 45 percent, and our Opex grew at 21 percent. We're already seeing leverage in the model. It's all about, as a company, finding the balance between adding to the expense line, particularly on the sales and marketing side, and also getting to profitability. A few years ago, the market valued growth at all costs. Today, the market actually values a path to profitability, and that's what we're doing.

I can't make forward-looking statements. But I can comment about the past. And we're already reducing our expense growth relevant to what our revenue growth is. We're very confident that we're on the right path.

As a public company, Tintri will now have to talk about how it will get to profitability and make investors happy. What's the long-term plan to do that?

I can't talk about what the time frames are, as I can't make forward-looking statements, but obviously we have an internal plan to do that. The basis of what we're doing is really three things. One is increasing our presence in large enterprise customers. That's one of the things that distinguishes us. …

The second thing, which is related to that, is improving sales force productivity. And a lot of that is about more channel leverage. And the third thing is increasing our gross margins, which is really around software.

That's what the plan is. There's obviously work to be done, but it's something we feel confident [about] given the quality of our customers, and given the trends that we saw in the last fiscal year.

What part of Tintri's business goes through indirect sales channels now?

We are a channel company. Our goal is to push as much as we can through the channel.

Is Tintri's enterprise business mainly direct or channel-driven?

It's primarily channel-driven. Our default is, we want to have channel partners involved. It's a very important part of what we do. That's a big focus of making sure we keep the channel happy and that [partners] benefit from our growth.

Is Tintri in channel partner recruitment mode?

We are in channel recruitment mode. … We definitely have an approach of focusing on partners that are prepared to invest in us, and not just darkening the skies with channel partners. We want them to make money as we make money, and we're absolutely not going to be overdistributed.

Do you see Tintri as some day being an acquisition target?

We don't comment about acquisitions. Our goal is to build a significant stand-alone business. Companies are not sold. They're bought. We're very confident, given our plan, given the 19X repurchase numbers, and our gross margins, the Opex leverage that we've seen in the last year, that we can be an independent company.

There was speculation that the Tintri IPO was more like another round of funding than it was a traditional IPO, given the need to build up cash to execute on long-term strategies. Any truth to that?

We could have done a private round. But we had alternatives in terms of doing an IPO. Our venture capitalists have been extremely supportive. However, given the nature of our customers who are large enterprises, most of them are public companies, they like dealing with public companies. We viewed that as being one of the most important things for us in determining that we were going to do an IPO.

What difference does that mean?

When you're private, people don't really know how big you are. You could be from $5 million in revenue to $200 million in revenue, and people kind of lump you in the same area. When your financials are public, they can see that you're not in the small numbers. And there's a marketing benefit from it. And what we have now is a fully funded model to get us to cash flow break-even and profitability.

So this was a very conscious decision. It is more important for us becoming public and not just getting the cash.

What are some things we can expect from Tintri going forward on the product side?

You'll see product announcements from us later this year. A lot of what we're doing is focused on the enterprise cloud, expanding beyond storage to compute or analytics, and additional self-service opportunities going beyond storage. And also increasing the scale of our systems. So today we support 160,000 virtual machines, which is one of the reasons that we're used by large enterprises. And we'll be expanding those numbers over time.

So, we're really focusing on large customers -- the Comcasts, Toyotas, Sonys, and so forth -- and doing the things that are important for us: Adding additional software support. And distancing ourselves from being defined just as a storage company.

Why is it so important for Tintri that it not be perceived as a storage company?

[Customers are] not interested in something that's just storage. We've seen over the last couple of years [that] market has really moved to us in that cloud is now the conversation that people want to have. I've talked quite a lot to channel partners and asked them the same question: Are you seeing this conversation changing around the cloud? And they say yes. It's a real opportunity for the channel to actually build a significant presence. That we have a cloud certification program that's specifically targeted at the channel. And this is a good services opportunity for the channel.

Can we expect Tintri to enter the hyper-converged infrastructure market?

Even hyper-converged people are kind of distancing themselves a bit by identifying in many cases as [providers of] enterprise cloud or private cloud. And I [believe] the approach we are taking to separate compute and storage is a more compelling approach. It's a more scalable approach. It allows customers to use their existing servers, and not lock [themselves] in. It doesn't require them to use your own hypervisor.

We believe that the big trend in the market today is cloud. It's obviously public clouds, but it's also enterprise cloud. … That's the telling opportunity. And we're very well positioned to take advantage of that. And again we have the proof points with the large enterprise customers. That's what we care about.