[In colloquial English, a “double whammy” is when something causes two problems at once.]

The media (and investment analysts) have focused on the positive news to come out of Autodesk’s Q2 earnings release, all showing a sequential improvement:

• Manufacturing division revenue up 4% over Q1
• Revenue from Asia up 3% over Q1
• Revenue from emerging economies up 6% sequentially, “led by low double-digit sequential growth in the emerging economies of Asia Pacific” (but emerging economies still account for only 15% of total revenue)
• 3D product revenue up 1%, “led by sales of Inventor, Moldflow, and Revit”
• Though revenue from the AEC business again declined in Q2, the rate of decline slowed from
-12% in Q1 to -2% in Q2.

We also learned that Autodesk’s channel performed better in Q2 than did direct sales (the opposite of Q1) and that Autodesk is not pushing inventory into its channel partners to boost its own sales performance, as the channel is now holding inventory for 3 weeks, down from 4 weeks. Autodesk also reports an increase in evaluation and pilot projects, although with fewer seats than in the past.

But there is still real cause for concern. License revenue in Q2 was $231 million, down 48% from last year, “due primarily to a 46 percent year-over-year decline in new seat license revenue.” Here are two of the three whammy elements: The decline in license revenue is due to the global recession depressing sales AND the company’s transitioning customers from upfront/perpetual to subscription-based license models. A subscription means that instead of recognizing $5,000 all at once from a single-seat sale, Autodesk can only recognize a small portion ($1,000?) as license revenue; the rest is maintenance. So in the absence of a lousy economy, license revenue goes down but maintenance revenue goes UP as each subscription kicks in.

But instead, maintenance revenue has been flat at $180-$186 million for the last five quarters. Q2 revenue was $184 million, up 3% over last year and up 1% sequentially. These increases are likely due to the change in license method, since “maintenance billings continue to decline due to a decrease in renewal rates as well as fewer new seats sold. Renewal rates have been impacted as customers have reduced their work force resulting in fewer seats to renew. “ In all, it seems the total number of seats under maintenance went down by 1,000. That’s not good — and it’s a spiral that will keep going downward. So this is really a triple whammy: lower license revenue and maintenance revenue because of the economy and lower license revenue due to the planned change in business model.

The real question is whether license and maintenance revenue can ever rebound to the levels seen before the business model change and the recession. Probably not: Manufacturing and AEC customers would need to add more employees and therefore more seats — something totally outside Autodesk’s control and on a timetable no one can predict. It is unlikely that the automotive industry will ever reach the headcount it had before all of the restructuring, so Autodesk is positioning itself (as it every other maker of engineering software) to enter new verticals and to expand into new application areas such as CAE within its core verticals.

Autodesks guidance for Q3 of revenue between $400 million and $420 million would be flat to a sequential decline and a decline of 32% from a year ago.

The good news is that Autodesk’s expense cuts have been keeping the company in the black and allowing it to add cash to cushion further downturns. But it will be interesting to track the maintenance numbers going forward and see how they lag the new license sales data. How low can maintenance go?