The 'Brochureware' Era Is Over
Hardware and software vendors need to change sales strategies.
The tech marketing industry is weird, and buyers' responses equally so. Very rarely are the products that actually sell in volume the models that are actually advertised, offered to reviewers or demoed at events. Feeds and speeds are cherry-picked, prices very carefully never mentioned, and vast feature gaps between demoed goods and purchased models swept under the rug.
'Technology You'll Only Ever See in a Brochure'
I recently heard a term for this phenomenon: "brochureware." The wikitionary definition of brochureware is "promotional material converted directly from existing printed brochures into a website, and thus lacking interactive features." I think the this traditional definition makes a great double entendre for the more modern, "Technology you'll only ever see in a brochure."
Both versions scream "You're doing it wrong!"
As a fairly well-known example, let's take Intel's 2-socket (2600) Xeon CPUs. The top end model sells for more than $4,000 US. Intel doesn't ship a lot of these. What it does sell a lot of are CPUs at the $1,500 and $1,000 per-chip price points. Despite this, you don't see a lot of CPUs from these mass shipping models in demo booths or on reviewers' benches.
As a reviewer, I usually get $3,000-ish CPUs in the servers that cross my lab. Not the top of the top, but also completely disconnected from what my readers are actually going to see in their systems. Looking at the CPUs that public cloud providers seem to be populating their equipment with, what they're buying isn't $3,000 a socket, either.
The storage industry is just as bad. How many 4PB+ All Flash racks of mind-blowing super storage with IOPS numbers so high they're actually meaningless do you think all the vendors littering VMworld's Solutions Exchange actually sell? But it's the 4PB All Flash rack they haul out to the event; not the 200TB hybrid array that they actually end up selling.
The Problem of Pandering
Brochureware isn't just about selling someone a 400-horsepower drag racer for their morning commute. There are plenty of examples of vendors pandering to the mass market with a crippled "SMB" model that they don't actually intend to sell. So many things are purposefully wrong with it, that the only point of the exercise is to be able to advertise a price similar to competitors -- who are actually succeeding -- into that space so that their sales team can try to upsell the customer to a much more expensive model.
While this has been standard practice for quite some time, the brochureware approach is starting to have deleterious effects on customer outcomes and vendor reputation. The reason this has become an issue of late is the swing away from discrete components and toward joined-up solutions.
At a hardware level, we're moving back towards great big, highly-integrated CPUs with graphics, crypto, neural networking and who knows what else, all integrated into one die. On the systems side, hyper-convergence, hybrid cloud and so forth are taking over. This sees dedicated storage, WAN accelerators, some layers of networking and more all folded into the server.
"Speeds and feeds" are no longer enough. We aren't buying a) IOPS and b) networking bandwidth. We're buying carefully balanced systems where a change in X impacts the availability of Y; and that cascades to alter the peak performance of Z.
With converged technologies, whether at the hardware or at the system level, seeing the model of gear we're most likely to buy actually matters.
Changing Consumption Needs Changing Marketing
IT is moving away from an area of simply overbuying various nameplate specifications and then slowly consuming them over the course of a refresh cycle. We're increasingly buying holistic, balanced systems that take expertise, experience and careful modeling to understand; and we tend to expand these systems on an as-needed basis. As our consumption needs have changed, so to must the selling habits of vendors.
Trevor Pott is a full-time nerd from Edmonton, Alberta, Canada. He splits his time between systems administration, technology writing, and consulting. As a consultant he helps Silicon Valley startups better understand systems administrators and how to sell to them.