Everybody wants a bargain. But when it comes to the complex world of IT products, finding a deal or even knowing what something should cost can be tricky. According to a recent Network World poll, 62% of IT buyers say the pricing structures used by enterprise network vendors are “usually confusing” or “very confusing”.
However, Dave Leonard, chief technology officer at outsourcer Infocrossing, has figured it all out. With extensive experience as a buyer and seller, Leonard knows all the tricks for getting discounts from vendors.
There are basically three things that motivate a typical vendor, Leonard says. The first is obvious: the vendor want new sales. The second is less obvious: the vendor aims to displace competitors, and may even set aside “displacement funds” specifically to give discounts to customers who agree to get rid of a competitor’s product. The third driver is the fear of losing ongoing revenue streams from maintenance and support costs.
Customers can use this knowledge to get better deals, even in non-competitive markets. Salespeople want to close deals before the end of the quarter because they are under constant pressure to meet goals for each three-month period. In other words, make a sales representative sweat for a few extra weeks towards the end of a quarter and you might get a discount.
“Even if there’s not a competitive situation, using time against the vendor gives them the opportunity to sweeten the deal,” Leonard says.
Leonard pulled out all the stops recently when Infocrossing embarked on a standardisation initiative across four datacentres. The project saved the company US$14 million (NZ$20 million) through consolidation of labour, software and other costs.
Infocrossing, which operates in 12 US states, has quadrupled in size through three major acquisitions over the past four years. The company ended up with datacentres running three different server management tools, from CA, IBM’s Tivoli division, and NetIQ.
Infocrossing decided to standardise on NetIQ after an evaluation of the products, but didn’t tell the three vendors that the decision had already been made. The first step was to build a business case showing each vendor how much it would cost internally to use their products.
“We did a complete economic analysis to get to ‘what will it take us to get to a single platform,’” Leonard says. “The cost is kind of what we presented back to the vendors. ... The idea behind that was to get them to understand that our cost of using their product was far greater than the actual cost the product was going to be.”
(For further exploration of this theme, See John Fontana’s article “Calculating TCO properly is Vital” on page 24).
Leonard told the vendors that Infocrossing didn’t want to be flooded with consultants, because its own employees would have to run the system. “We did say you can help the overall economic case by affecting how much maintenance we pay on our existing install base,” Leonard says.
He also asked vendors to loosen restrictions on existing contracts, such as clauses that prevent a product licence from being used in more than one datacentre.
Each time a vendor offered a proposal to entice Infocrossing, the company was able to bounce the idea off the other two vendors and ask them to do better. “We’re trying to end up with something that’s defensible on both sides, because they have to sell it internally,” he says.
After a negotiation period of three months, Infocrossing got a deal from NetIQ that Leonard says will save the company “seven digits” over the next five years.
The company was already running NetIQ on about 1,500 servers and wanted to standardise across 5,000. After the wheeling and dealing, the licence charges for the additional 3,500 servers were “negligible” because NetIQ funded the cost with competitive displacement money.
“It went from a significant to an insignificant cost of the whole operation,” Leonard says.
Infocrossing also focused on maintenance costs, because paying 20% of list price, as vendors would prefer, “can just kill you”, he says.
Leonard’s goal is typically to pay 20% of the acquisition price, and negotiate clauses that limit price increases related to future acquisition of licences. Immature IT buyers often make the mistake of focusing only on upfront costs, when future costs for maintenance and additional licence acquisitions can turn a seemingly good deal into a bad one, he says.
The process Leonard used to negotiate lower server management costs was replicated across 20 or 30 products in the standardisation initiative, making software a significant portion of the cost savings achieved in the whole project.
Infocrossing still uses many IBM and CA products in areas other than server management. The company ended up paying NetIQ more overall than it did previously, but the cost per-server is “way less”, Leonard says. He can’t say exactly how much they paid due to a non-disclosure agreement.
“If they don’t hamstring you with a non-disclosure agreement, that generally means you didn’t get good pricing,” Leonard says. “They don’t want our pricing available to the general public.”
Infocrossing often finds itself on the other side of the table, when its own customers ask for discounts.
“Sometimes, we’ll say ‘absolutely’”, because when Infocrossing’s hardware costs go down, it makes sense to pass some savings on to customers, Leonard says.
“If there’s not a basis where our costs have gone down, we go back to them and say ‘hey, here’s what our costs are, there isn’t anything that’s changed since we did the deal before. It was a good deal then and it’s still a good deal now,’” Leonard says.
Leonard says a vendor that is logical and unemotional can typically convince a customer that the price is right, even if the customer has asked for a discount.
“The customers don’t know, largely, how much things should cost,” he says. “The more confidence they get from us that we know how much things should cost”, the more confidence they will have in the pricing, he says.