Intel reported solid earnings last month, and we don't need to go into details here; instead, we want to focus on one particular aspect of its data -- long-term gross margin guidance. In the most recent quarter, they reported a gross margin of 42.5%, but importantly, they reiterated their goal of one day reaching 60% gross margin.
Six months ago, most people, including us, thought this goal was a fantasy at best. But now, we have to at least consider the possibility, however remote, that they might actually pull it off. In particular, it's critical to understand how Intel is restructuring internally in preparation for the launch of Intel Foundry Services (IFS) as a third-party foundry for semiconductor designers.
Earlier this year, Intel announced changes to its internal accounting methods. Under the new system, Intel's product groups and factory operations will each have separate profit and loss statements and interact on an arm's length basis. Previously, costs from the two divisions were blended together, with operating costs hidden in product margins.
While this may seem like a minor accounting adjustment -- after all, the company will still report consolidated gross margin, which will remain the same -- this subtle change will have a significant impact on internal incentives. Our core view on Intel is that their biggest challenge is changing their internal culture, so this "small" change could have a huge impact.
Our core view on Intel is that their biggest challenge is changing their internal culture, so this "small" change could have a huge impact.
Following the earnings release, we had a chance to speak with management about their performance and dive into how they plan to achieve their long-term goals. One important takeaway from this conversation is that management claims they are implementing a number of "simple" internal measures to improve gross margin.
Some measures are straightforward, such as comparing fab operating costs to industry peers. There are also measures that sound simpler than they are, such as charging product vendors for rush orders or "hotlots." The obvious question is: If these are low-hanging fruits, why haven't they been implemented yet? The answer is that these changes are not as easy as they seem.
Imagine you are an Intel salesperson responsible for a super reseller account (a major provider of cloud computing, networking, or data storage) whose data centers are major consumers of Intel chips. Customers are about to make a major purchasing decision between Intel and AMD CPUs, and the order value may be as high as $1 billion. If you win, you'll receive a Cadillac and a huge bonus. AMD launched a new chip earlier this year, but Intel's product has only just arrived.
The customer wanted to test both options under real workloads in a 1,000-CPU system. Your product is in short supply, but using last generation Intel CPUs will likely give AMD the win. Therefore, you ask the operations team to urgently order 1,000 parts.
The operations staff are reluctant to do this because the product is new and the yield is low, so they have to produce double the number of wafers to get 1,000 good chips. In addition, the customer requires a specific SKU or version of the chip, but the operations team is ready to produce a different SKU. In order to get these 1,000 parts, they need to shut down the factory for a few hours, retool, produce chips, then close the factory again and retool as originally planned.
If the factory costs $30 billion to build, half a day of downtime could cost $10 million in depreciation. Under the old model, customers generally didn't care because you knew you were going to win $1 billion in orders, and all those fees were going to be rounding errors in sales.
Under the new model, salespeople will now have to bear the full loss of these expenses. Let's say the order was placed near the end of the year, so the expense would have to be incurred on this year's numbers and the purchase order wouldn't be available until the following year. Selling in this P&L situation will most likely get fired and someone else will benefit from the purchase order next year. However, if you don’t place a rush order, the deal may not work at all. What would you do in this situation?
Of course, this is a highly simplified scenario, but it aptly illustrates the profound cultural shift that's about to take place within Intel. Intel's previous approach has been integrated into its business model and sales strategy, giving it a significant advantage over its competitors. Without this tool, or crutch, sales teams will need to rely more on product performance itself.
This shift could affect its market share and revenue prospects in complex ways. As our hypothetical situation shows, a lot depends on individual decisions, team dynamics, and management culture. Does the salesperson’s boss think about the long term? Can the operations team maintain some flexibility? How will senior management handle disputes? Will management agree to a one-time exception and then make it the norm, thus defeating the purpose of the change?
Accounting can be boring, but it can also sometimes lead to very interesting developments in the actual running of a company.
Editor's note: Guest author Jonathan Goldberg is the founder of D2DAdvisory, a multi-functional consulting company. Jonathan has developed growth strategies and alliances for companies in the mobile, web, gaming and software industries.