Project Blowfish: Look like the brand you want to be

We like what you guys are doing but there’s no way we’d replace the aorta of our communications infrastructure with a beta box from a 20-person startup” – Director of IT, Fidelity Investments (circa 2002)

Call it the enterprise startup conundrum: how do you earn legitimacy if no one will give you an opportunity to become legitimate? The enterprise world is different from the consumer world: barriers to entry are higher, switching costs away from incumbents are bigger, and the risks for failure (for both buyers and sellers) are greater. And while there’s definitely no single right answer, an enterprise startup that can’t crack this conundrum is a dead enterprise startup.


At IronPort, we had built a radically better email gateway. It crushed the popular and arcane sendmail — the free, open-source software that we aimed to replace — in every dimension. Our product was even cheaper to own, as it ran on one-tenth the hardware, but we were unproven and email was mission critical. To make matters worse, for our early adopter industries (ISPs, media, tech, and financial), no email service meant no work. We were so good, yet so puny.

We felt like the little kid that kept getting turned away at the height chart at the rollercoaster — we can handle the ride, just give us a chance! Somehow, we had to find a way to look bigger and more credible quickly. After an intense brainstorming session, we hit upon a really important concept: since perception was reality, any weakness that the customer couldn’t see and couldn’t touch did not exist. By being absolutely maniacal about each of our customer touch points, we would appear to be far bigger than we were. No matter how rinky-dink the “man behind the curtain” was, it didn’t matter because the Great Oz would be massively impressive. We called it “Project Blowfish.”

We examined everything the customer could possibly touch. It had to look like it came from the big boys: Cisco, Microsoft or Apple. How would they do it? We asked and researched. Here’s what we came up with:

  • We looked big. We paid up for the best graphic designers for the website, business cards, and the sales collateral. Heavy cardstock business cards. High gloss datasheets and well-written content. Smart, insightful— as perfect as possible.
  • We felt big. Our appliance that sold for $ 50,000 was at its core a cheap-looking Dell 2650 2U rack-mounted server running our software that cost $ 3,500. At great debate from our board, we hired IDEO (www.ideo.com) to design a thick, aluminum faceplate and backplate with our branding. And they did it for equity. (Here’s a picture: http://www.ironportstore.com/web-security.asp). The shipping box was packed smartly with very high quality cardboard that had our logo silkscreened all over it. It was reminiscent of how Apple packs it products.
  • We projected total professionalism. For the physical manual, we took inspiration from the Sony Handycam and the Cisco Router manuals. It had clear diagrams, was well organized, and had a laminated Quick Setup guide. We shipped all of our evaluations out via Fedex next day, which made a really strong impression on customers who typically waited weeks for competitor’s products. They got it much faster than expected and when they opened it up -everything they saw was impressive.
  • Our ads were on brand as well. We took out full-page ads in major IT publications. Great product pictures, clear, clean, and uncluttered messaging.

All of these details mattered, but the key to the entire plan to make our little company appear big was to perform like the very best large companies. For us, this meant dramatically beefing up customer care (CC). In general and especially in high tech, CC had been a backwater — treated like a cost center to be minimized. Since this was arguably the most important early customer touch point, we took a completely opposite approach: we treated it like a marketing expense. With a firm belief that this was the primary catalyst for word of mouth, we invested heavily in the entire post-sale ecosystem. Our approach was to answer email inquiries instantly and telephone calls on the second ring with little to no hold time. We concentrated on how to deliver the best possible experience first and then went back and figured out how to cost-optimize it later. Here are a few ideas that ended up working:

  •  For the first six months (until we were absolutely overloaded), the entire leadership team was on the CC email alias. We were aware of every issue, saw the thoroughness and timeliness of the responses and chimed in with suggestions. This helped set the customer care culture: it was damn important and always had a bright light on it.
  • We hired the best of the best to lead CC and gave them a ton of autonomy to make decisions quickly. (I’m extremely proud that our first two heads of CC subsequently became venture-backed CEOs: Cyan Banister and Patrick Peterson.)
  • We hired the best all-around athletes, paid them above market, trained the hell out of them and then gave them a legit career path into sales engineer, engineer or CC management. We hired for friendly, intelligent and responsive…
  • We tracked all of the metrics relentlessly and then posted them everywhere. Big screen monitors hung in CC and showed constantly moving progress charts which then linked to internal status websites and dashboards. We had green signal lights around the office that turned yellow and red when a customer was on hold too long. The visibility throughout the company helped to make it a special place to work.
  • We opened in different time zones early for 24-hour coverage. We found that the best people would rarely work the graveyard shift so it was much better to hire in a time zone when they were naturally awake. I remember this saved our bacon during our evaluation at Goldman Sachs. Goldman, based in NYC, had a CC acid test for hot tech companies, especially ones from California. They’d send in an urgent email at 7am EST or 4am PST. Our newly opened CC center in London picked it up in 30 seconds and gave a detailed fix and thorough answer. They were floored.
  • We had a direct connection between CC and product management with regular reviews. In fact, some of our best ideas came from customers working through problems with a CC rep.

Ironically, the most fanatically apostle customers were the ones that had the biggest problems—catastrophic failures. It was during those times that we showed our mettle. We had a funny saying about it: “When you save someone from drowning, they quickly forget that you were the one that pushed them into the pool in the first place…” In retrospect, by focusing on all of the touch points in an effort to look bigger, we actually became that company. What started out as a bit of a façade became real and we grew into the oversized shoes.

Scott Weiss is a general partner at venture capital firm Andreessen Horowitz. He is the former co-founder and CEO of IronPort networks, which was acquired by Cisco in 2007.

Image courtesy of Flickr user pterjan.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

  • Flash analysis: lessons from Solyndra’s fall
  • Flash analysis: the tech startup investment environment, Q3 2011
  • Millennials in the enterprise, part 2: benchmarking IT’s readiness for the new digital workforce



GigaOM