Pre-Announcing as a Strategy

Pre-announcing a product might seems dumb, yet it happens often. Is there some reason this keeps happening?

Steven Sinofsky
Learning By Shipping

--

Source: clipart

Oh no, Apple finally announced AirPower would not ship and even apologized, on a Friday no less. Pre-announcing a product comes with an inherent risk that sometime down the road you will face the possibility of announcing that a product won’t be released, so why do companies end up in this situation? As with most topics in business, there’s no clear-cut answer and the context matters. Pre-announcing can have strategic value even if a company fails to deliver.

The idea of pre-announcing a product at first seems really dumb. When you think about an action like this you often come to a conclusion that anything that seems dumb but happens often probably has some clever, and strategic, use for it. Pre-announcements are no different.

In today’s world of constant delivery of SaaS “customer value” for enterprises on the one end and cool viral adoption for consumer technology on the other, we might have lost sight collectively of the value of “product launches.” A product launch is a unique opportunity where customers go from no knowledge about a product to the ability to use (and own) the product in one swoop.

In many ways, Apple has come to define the best of class execution at captivating announcements coupled with near instant availability. In fact, Apple is generally considered so good at this that people often forget just how awful Apple once was or even what deltas we see today. After Steve Jobs became return CEO Apple went through a period where new products were announced and availability, immediately or otherwise was often so extremely limited that it was effectively unavailable. This was one of the things Tim Cook and the operations team fixed about Apple.

Still even has the company matured, there were times when products were immediately available and times when they available in a short time later, and times when availability was far off. While we remember a product being immediately available to have or to order, in practice availability for announced iOS products has averaged about 4 weeks. There were some shortages and some outliers such as the original iPhone announcement in January and availability on the last day of June or the iPhone 6 availability immediately. By and large though, people have rightfully come to expect Apple to announce and immediately deliver. And the announce would be a surprise and delivery would be certain.

This is so baked into how we think, that I recall being given a hard time by a well-known Apple blogger/writer when we announced Surface PCs in June along with availability in October. At the time it was written that Apple would have shipped them that week, which was never true in reality then but the perception was there. Yet Microsoft’s track record was probably the exact opposite (especially at that that time, given that the most current release of Windows shipped in 2007 after being announced in 2001 as shipping in 2003.) The good news for us was that for that announcement we landed the date (two dates actually) as announced, as planned.

We don’t know reasons behind Apple’s variances. I have no idea. I certainly have no idea what was going on with AirPower. I have no interest in speculating either.

I am, however, interested in the concept of pre-announcing and why or why not it might be a good idea. Whether the decisions involve consumer products or enterprise, there can be a strategy at play. It just depends on the business.

Disclaimer, sort of: For many years I worked on Office and developed a reputation for never pre-announcing or detailing “future products”. That was true and there was a reason. I worked on Windows then and as part of the job we had to announce things long before they were finished. This post reflects that experience.

In many products the idea of never announcing until you are done is a way to maximize the bang for the buck — it generates massive excitement if done right. This is how people sometimes perceive Apple (even though there was that six months with the iPhone). If customers can and want to act immediately then having no announcement is spectacularly compelling.

In practice, things that costs $1000 or purchases like enterprise software that go through a big corporate process don’t really adhere to these rules of consumer excitement. People and/or organizations will use time to weigh alternatives or test suitability to products. Surprising the market with a product is fun for the team, but to customers, especially in business, it is sort of a headache. This is especially true if a product comes out at some arbitrary time seemingly for no reason.

Still, there are several reasons why pre-announcements might be a part of a product or business. These might not be “good” reasons but they are, nonetheless, reasons:

  • Early announcement, but not finished. First and foremost, a product might not be finished and what seems like a pre-announcement is really just an early announcement. If a team has particularly good at execution or have enough buffer in a schedule, this might be routine. It might seem to the market that this is a pre-announcement but in practice this is just announcing before availability.
  • Competitive. A product might be neck and neck in a competitive battle and the primary competitor either announces or rumors swirl that it will announce a feature. As a result, the choice is made to pre-announce in an effort to draw attention to efforts so that the business won’t be viewed as behind or potentially, worse, described as “catching up”. Believe it or not, when Microsoft pre-announced touch support in Windows 7 in May of 2008 it was due in part to the rumors of an impending tablet from Apple (which we of course thought would be a Mac).
  • Rounding out an offering. Often as a product is maturing and going through the growth stages, it becomes clear that customer after customer is blocked on closing deals because something is viewed as missing or the product is incomplete in some way. Often this is something like extensibility or a connector to a leading but adjacent product, or maybe some sort of migration/conversion tool to move from the product being displaced. These pre-announcements are often driven by sales and marketing. The view is that the announcement will serve to accelerate deals or at least create the belief that the product is moving forward and objections are being handled. As you can imagine this is very difficult to pull off in a practical sense since if the features are truly blockers the customer is still blocked. In enterprise software, however, many things that are described as blockers are more about negotiating than anything else. Maybe this was AirPower?
  • “PR”. Sometimes a pre-announcement is literally intended to create a communications or public relations pop. By this I mean the announcement is done out of band and unrelated to any specific event. It is done for the value of making an announcement. In particular, this is aimed at generating buzz or raising awareness. Sometimes these announcements are made under the guise of market feedback, testing the waters, or part of a broader launch plan. You can spot these announcements because details are generally fairly scant and in particular there doesn’t seem to be a time frame on when the announcement will be actionable. But who doesn’t love a good viral campaign? Well, you’d just better deliver.
  • Big event. Sometimes a company has a big event and is looking for news or something exciting to say. This is particularly the case for a big company where there might be an annual conference (or even big internal sales kick off) and the exec leadership looks around and just doesn’t feel there is enough exciting to talk about. In this industry, I would say the vast amount of stuff that is pre-announced falls into this category. It is amazing how much pressure a CEO keynote places on product roadmaps.
  • Leak. Pre-announcements are not always planned events. Sometimes when a leak happens it is not possible, or credible, to avoid comment. Eventually the information about futures piles up and an announcement is made sort of under duress. In enterprise software a leak eventually leads to a big event as above when the product still isn’t ready.

Many of these have one thing in common in that they are viewed internally as a forcing-function. That is the act of pre-announcing acts to force the hand of the product and business team to alter plans and now deal with the new world order as pre-announced. In general, almost nothing good ever comes from a forcing function and so I would say that once someone feels that pre-announcing is a tool used to drive change within a company, that is a likely point of failure. I will spare you from many examples where something I was working on was pre-announced as a way of getting commitment :-(

What I describe above are times when you might want to pre-announce. In these cases, pre-announcing is a strategic (or tactical) approach to product launches. The hard-core business requirement isn’t there. It’s just a choice.

There are times, however, when pre-announcing is not only a legitimate part of the business but is actually required by the way a business operates.

  • Marshaling partners or ecosystem. Many products are acquired along with a network of supporting products and partners that operate as distinct businesses. Products that can only be sold as part of such an ecosystem will often be pre-announced such that those independent businesses have time to plan, build offerings and technology to support a product, and get any supply chain spun up. As a company grows, much of an ecosystem often develops that can be partnered with behind the scenes, but early on when building out an ecosystem and attracting new partners a pre-announcement can be a sign of commitment to the marketplace. Maybe this is why a company might pre-announce a subscription game or streaming service?
  • Sales, Support, and Distribution. Many products require time to get to store shelves or in the hands of partners. Of course SaaS software doesn’t require that, but it might require training and support to partners as part of the selling process. In this case a pre-announcement is required in order for the work of stockage and training to take place. Here again these independent businesses can’t just change on a dime. A classic example of this is stocking for holiday sales — most businesses plan retail holiday shelves and floor layouts nine months in advance. While much of this can be done behind an NDA, again partners will often look to a pre-announce as a strategic commitment or “skin in the game” on the part of a business.
  • Customers demand lead time. In enterprise software customers demand lead time. A pre-announce is a way of maintaining a strategic relationship with customers that have their own strategies for a product category and in order for a product to be considered customers need time to build that into planning and strategy, even if they are not buying right away. IBM mastered this type of approach to selling and while it is easy to see the “evil” side of this it is the norm in enterprise IT.
  • Signal a whole new business or expand the view of a company. If a company is making a major business shift that takes it into a new market or takes an existing product into a major new business then pre-announcing is a way of laying the groundwork. Often a whole new business for a company means connecting the dots across ecosystem, sales & distribution, and the need for customer lead time. A whole new business has the sum of all the needs for pre-announcing.

The challenge of pre-announcing when the business requires it, is that there’s nothing in there that says execution must follow. In other words, the need to pre-announce does not automatically imply follow through on execution.

That means it is entirely possible that a required pre-announcement will down the road be followed by a humble apology or perhaps just a new announcement without any recognition of what might have been previously announced.

The PC industry is haunted by two legendary stories of pre-announcements. These contribute to fears of pre-announcing or at the very least labels for critics of pre-announcement to use.

The Osborne effect is when a successor product is announced and as a result demand for the current in-market product dries up. The result is a precipitous drop in sales followed by a cash flow crunch and an inability to launch the new product that was just pre-announced. Our family business had an Osborne I and when it came time to buy a second machine we actually postponed buying one because the new model was announced. We never got a second Osborne before IBM PCs took over. This is a very real risk that must be considered. It is especially as problem for physical goods/hardware or for products that are selling extremely well and are not amenable to replacement purchase.

Vaporware is a phrase from the early 1980’s describing products that are announced and never existed. It is often described in a negative way referring to products that were announced in an effort to freeze out sales of competitive products. In reality, this phrase came about at a time when actually finishing a product and releasing one in a state that worked for customers was a huge task and often simply failed. Most of the famous vaporware products were not evil, but simply got in a state of too many bugs, used too much memory, or fell behind a competitive product that was actually released.

Historically and my own experience (I promised a story) Office and Windows demonstrate both ends of the pre-announce spectrum.

The Office business (and all of productivity software) grew up originally as a retail business and one where hitting an actual target ship date almost never happened. As a result, the desire to drive people to the store once and to capture their business that time was critical so there were few pre-announcements. Any pre-announcement would have immediate impact on sales of the current version. Over time the industry developed “upgrade” pricing and other practices to encourage more frequent buying but still would not routinely pre-announce, except as a specific strategy (see above).

As the whole industry evolved to selling large volume multi-year (pre-SaaS) agreements to business, the problem still persisted because even though these were annuity agreements the business still owned outright the software. Pre-announcing would have the effect of slowing down deals (why close now when the new version is soon). The industry also matured at delivery and product shipping became more predictable. While pre-announcing still had substantial negatives a mid-point was reached where under NDA and in smaller sessions the “strategic direction” of the next release would be shared. It was sort of a pre-announce.

SaaS totally fixed all of this, or so we think. Today there is almost no upside or downside to pre-announcing when it comes to deals since once you have a customer you have them for what actually works right then and there—customers bought the product for what it does right now and the stability that comes with that. Still the desire for a company to show off this value subjects us all to popups, emails, and overlays alerting us end-users to amazing new capabilities that were all once the domain of launch events and trade coverage.

Today’s reality is that customers are becoming more annoyed with changes to business SaaS products than they feel value and would like to return to a more regularly scheduled delivery cadence with pre-announcements. No one likes a morning surprise of “who moved my cheese” in a product for getting work done. On the other hand, if we think quarterly updates are what customers want, then consider Patch Tuesday. For expanding/growth products frequent updates are fine (product market fit assures us of that). I would say no one has really figured out the right way to deliver either upgrades or launches for SaaS in a customer and purchaser-friendly manner.

The Windows business contrasts sharply with the Office business in that it depends immensely on a broad ecosystem, hardware partners, and distribution for sales. This requirement means Windows has perennially been in a state of pre-announcing products, literally since the very first version of Windows that was hastily announced at a COMDEX trade show and not delivered for two years (after it had been under development for 2 years prior). Why would anyone do that?

Creating an operating system that requires third-party PCs, peripherals, and applications cannot be done in a vacuum. All of those independent companies are presumably doing other things already (like making printers for Apple ][, or a word processor for CP/M, or a bios for one of dozens of other PC companies). In order to get those companies to make plans to work with Windows, Microsoft had to be way ahead of reality in marshaling support. Imagine trying to get a company in Japan to commit resources to building a whole new BIOS, disk controller, and software for a PC before an operating system exists? That’s what it was like to get the PC clone business going, especially early on when there were so many other ways to use resources to make a living because the PC had so many competitors.

As the industry evolved this dynamic became even more difficult and important. PC companies became huge corporations with their own processes and priorities. It wasn’t just planning for a holiday sales season, but it was planning across an entire supply chain for that holiday and doing so with a huge focus on COGS, demand generation, SKU mix, and more. The hardware requirements, new peripherals required, device drivers, and more add up to an immense challenge. PCs in the growth stage were highly cyclical since they were big holiday, back to school, or fiscal year purchase items. How well a new version of Windows or a new Intel chip landed with support for any major new technology would make or break a PC model (DVD burner, USB, Wi-Fi, etc.) which would make or break a whole season and investments in all of this.

I have described a number of reason why it may be necessary to pre-announce. Recall though that pre-announcing does not imply flawless execution. Reading all these reasons though, one can see the immense cost of pre-announcing as a strategy when delivery does not support it. That is what happened over the course of many years with Windows and was so difficult — the trust across the industry can become lost and need rebuilding. Product-market fit is a remarkable thing given that the pre-announcements did not slow the demand for PCs or ability to fulfill demand, but failing to deliver on pre-announcements eroded trust over time.

Many people can often become confused by the difference in how much a company shared across different businesses or even the culture of information sharing. Some people may believe this was a matter of choice or personality, when in fact it is driven by the varying needs of the business.

Pre-announcement is a strategic tool that can and should be employed in the right circumstances. Just have your execution lined up to support it.

— Steven Sinofsky (@stevesi)

--

--