Original URL: https://www.theregister.com/2002/09/25/the_red_hat_ibm_pitch/

The Red Hat, IBM pitch for Linux in the enterprise

It's the salespeople, stupid...

By John Lettice

Posted in Software, 25th September 2002 17:01 GMT

Earlier this week The Register had an interesting talk over lunch with Red Hat VP marketing Mark de Visser, covering a fair range of topics, including how Redmond finally gets it, what takes its place (no, not Red Hat, we agreed on that) and the two Next Big Things for Red Hat - the Red Hat Desktop, and the Advanced Server deal with IBM. There's a certain amount of controversy associated with both of these, but as (with the aid of the Red Hat 8 CD he pressed into our hands) we'll do a user's eye view of the desktop later, we'll concentrate on the Advanced Server strategy here.

Red Hat has been involved with IBM for some considerable time, IBM has been talking the Linux talk for some considerable time, but it is the perception of The Register (which in its turn, has been hanging around IBM for a not inconsiderable time) that much smoke and little light has emanated from the Blue Behemoth. IBM can be a great place to hunt around for Linux information and support, and it clearly has numerous happy techies messing around with the stuff, but how easily can you actually buy Linux machines from IBM? Sure, you can get them if you specifically ask for them, but the sales operation tends not to evangelise them.

This applies in one way at the client level, and in another at servers. Wintel inertia is epidemic at client, and does apply to some extent to the Intel end of servers, but upscale of this at IBM the server portfolio is sufficiently broad for this not to be the reason Linux doesn't get pushed. The reason, to precis de Visser brutally, is because it's free. Or it was, anyway.

IBM is a huge organisation with many most excellent salespeople selling large and expensive systems into areas where the word Windows is seldom uttered. The competition is Sun, maybe HP (but see our Mr Orlowski today on the latter's exercise in self-immolation) and to a very great extent, IBM itself. IBM's salespeople, as de Visser points out, sell the stuff that makes them money personally. They are, after all, salespeople.

So simply sticking Linux onto the checklist is not of itself going to help much. It will get Linux installed if that's what the customer specifically wants and asks for, but otherwise if the salesperson can make more money recommending something else, then that's precisely what they'll do. Evangelism failure. You want evangelism from sales? Bribe them.

The Advanced Server announcement foregrounds service and support and bringing the various IBM server offerings into step as far as Linux is concerned. These are certainly important, but the sales side is key. There's some sense in the argument that major companies are suspicious of software that's free, but making sure the sales team gets equal or better remuneration out of selling Linux is probably a bigger factor in getting it out there. De Visser is confident that with the IBM deal that is now the case, and that Red Hat now has working for it tens of thousands of feet on the street, all pounding in unison (Register rolls eyes incredulously at this point, but we'll see).

Pricing on Advanced Server is related to support level, and is an annual fee rather than an outright buy. That is however the kind of model this class of customer is used to. It is also still GPL and free, but Red Hat is supplying this in rigorous 'some assembly required' form - you want it for nothing, you build it out of the source yourself. The Register remarked that this was a sort of escalation of SuSE's approach, and de Visser brought up the question of whether we reckoned Red Hat could ship proprietory software. No, we said. Well what about SuSE?, he said. Yast is proprietory. Oh, we said dozily, we didn't know that.

Nevertheless, SuSE is different from Red Hat. Red Hat is the one the purists worry about, because it looks like the one with the clearest shot at hitting the corporate and government sweet spot, and in the act of doing so it could well turn into the kind of company that already occupies these sectors. So people are going to be ever more ready to be suspicious the more successful Red Hat is, and Red Hat has to be really good, only to be rewarded with deepening suspicion anyway, the more successful it is. Upside: at least Red Hat's rock and hard place aren't anything like as closely aligned here as Sun's. Yet.

Aside from getting more conventional and understandable to the non-evangelised in terms of sales and pricing, with Advanced Server Red Hat is also making a conscious effort to make it a single, coherent platform. As de Visser puts it, currently with Linux you'll get some distributions that one piece of software is certified for and runs on, but there will be other packages that won't, so if you've got two or three enterprise applications you want to run you might at least have to juggle different releases of the same distribution. This is not, by the way, forking as such. It is simply a case of Oracle might have implemented 9i at so and so level while A N Other company will maybe slightly ahead or slightly behind. It'l sort itself out in the end, but by that time we'll have had a couple more point releases, so it won't quite be sorted after all.

Therefore, argues de Visser, it is A Good Thing to go for a slower release cycle in order to get all of the key apps available at the same time on successive iterations. Which again makes sense from the perspective of the corporate market, but again will surely have the tendency to make Advanced Server a more distinct and, sort of, not quite Linux really product, as the rest of the market (Red Hat's other stuff included) goes charging off at full speed, not necessarily in rigorously close alignment.

The Register is nevertheless inclined to applaud Red Hat, on the basis that the route it's taking seems a logical way to gain commercial success for Linux in the world as it is now, whereas the alternative of overthrowing the world as it is now does not seem to us immediately viable. If something distinctly different is what it takes to crack the commercial market, then so be it, that beats not cracking it at all. As we said, neither we nor he think Red Hat will itself become The Beast (he might of course be lying), but even if it does we trust it will be a nicer class of Beast.

Should however the revolution turn into the directorate, the consulate and then the empire, you may come round and lynch us. You have our permission. ®