First-Mover Advantage or Winner’s Curse?

I decided to do a little research:-  the original idea came from a paper in 1988 by Lieberman and Montgomery which they followed up on 10 years later.  Then in 2005 Lieberman presented another paper reporting on new research he and his students had been conducting on whether it was a significant factor in the success of companies through the dot-com boom. I was particularly attracted to the last paper because I’d started to add “first-mover advantage” to my list of junk MBA concepts which do not apply to high tech, innovative or disruptive products.

Lieberman and Montgomery’s proposal is that there are 4 mechanisms which can sustain a first-mover advantage; (1) proprietary technology, (2) pre-emption of scarce resources, (3) customer switching costs, and (4) network effects.  You’ll recognise these as tricks we all attempt to use in order to raise barriers to competition, though some require a little explanation. 

“Pre-emption of scarce resources” is all about locking up key resources that your competitors would need; prime high street retail estate or access to scarce materials.  It is difficult to see how this relates to software unless you are lucky enough to come across IP which no one else is working on.  Sadly, my experience is that if one university department is working in an area, half a dozen will be, and none of them will have had the whit not to publish their results in detail, so making that scarce resource available to almost anyone.  Many software companies live in fear of “two guys in a garage” who could, by starting afresh and not being distracted by customers, catch up and overtake them quickly.  This is indicative of how proprietary their technology is and how scarce the resources required to create it are.

“Customer switching costs” concerns how difficult it is for a customer to switch from one product to another.  It could be the cost of implementation (anyone who has been involved in implementing enterprise wide software solutions will recognise this), the cost of training, or less subtle tricks like locking up all your data in a proprietary format.

“Network effects” arise “when the value of a product or service to a given user increases with the number of other users”.  It is pretty easy to see that Facebook, eBay and Wikipedia all benefit from this effect and make it really hard for others to compete.

What Lieberman described in 2005 was work to assess “the magnitude and sources of first-mover advantages in 35 Internet markets, based on data from more than 200 publicly traded entrants.”  Half way through the paper he admits that a “bias arises in cases where the market pioneer remained private or failed without issuing an IPO” (by the way, “failed” also includes being acquired; not a definition most VCs or entrepreneurs would accept).   So we really don’t know whether he is looking at “first-movers”, “fast-followers” or even “first-to-succeed” only “first-to-IPO”.  When you consider how long it takes some technologies to come to the mass market, “first-to-IPO” could be decades after “first-to-market”.

So what of their results?

They measured “proprietary technology” by measuring the number of patents.  Easy to measure, but in my mind neither a measure of the proprietary nature of the technology, nor as Lieberman suggests, the level of innovation of the organisation.  In the late 1990s I worked for Eastman Kodak.  A few seconds on the US Patent Office website suggests that they now have nearly 20,000 patents, yet my experience was that they would struggle to find an innovative place to have lunch and wasted huge amounts of time and resources by developing the wrong products.  Most technology start-ups I’ve worked with have two approaches to patents; (1) they don’t have the resources and don’t think the system works, so rely on physical security to keep their secrets, or (2) take out a small number of patents which they hope may they might be able to trade if a mammoth like Kodak comes after them and will add value when they exit.  None of them was silly enough to think they could actually afford to defend a patent.  I do believe that the number of patents does reflect a level of aggression, funding and intent by an organisation which should make competitors wary.  It is also worth noting that a patent may be more important to a “fast-follower” than a “first-mover”, after all, they need to carve out their own technological uniqueness and may need many more patents in order to work around the first-mover’s IP.  So, not much surprise that Lieberman found a link here.

He did not bother with “pre-emption of resources” as “we lack objective criteria to identify such markets” where this could have an effect.  The same is true of “switching effects” where “objective empirical measures are unavailable”.

As an aside, I believe that trying to erect “switching costs” is a very dangerous thing for a company to do.  If a customer recognises that you are intentionally, and unnecessarily, increasing his initial investment, reducing his flexibility, making him more reliant upon a single supplier (you), or failing to properly support technical standards, he is likely to take a dim view of your offering.  Alternatively you can erect switching costs by delivering exceptional service, offer a product that does just what the customer needs and more, and give them benefits that include quantifiable savings that deliver to their bottom line; then they’ll be happy not to switch.

As I pointed out earlier there are some business models where “network effects” have a big impact and investing to create this effect has to be a critical part of the strategy.  Lieberman admits that “it is not clear … how this … more general network effect can be effectively measured”.  For other markets it is far less clear.  In fact in some B2B markets where your product is a visible part of the service your customer offers their customers, it could have a negative effect.  Imagine standing outside two dry cleaners on the high street and trying to decide which to use.  Then you spot that they are both proudly displaying the “whizz-o-matic-clothes-cleaner”.  Suddenly the only factor you have to choose on is price; by selecting the same product they’ve made their service a commodity.

Lieberman concludes by finding a benefit to his “first-movers” in that they “had a market value roughly two or three times that of follower firms”, and that the “advantages have been minimal for pioneers that do not benefit from network effects or patents”.  I suspect that he would have found a similar link for any group of companies which were aggressive, well resourced and were able to achieve network effects, first-mover or not.

I’d like to propose an alternative explanation for what is happening here.  If these really are innovative or disruptive products, they are being launched into the unknown.  There will be no pre-existing market to measure, customers will not have bought similar products in the past, and may struggle to understand what these new ones do.  Just asking a customer about it will introduce what physicists call an Observer Effect (where the act of observation changes the phenomenon being observed).

In this environment the first-mover may be really, really, really lucky and get their product offering spot on first time.  More likely they will need to quickly adapt their offering as the early adopters figure out how they are going to use it.  How many times have you heard entrepreneurs say something along the lines of; “If I knew that we were going end up here I’d never of started from where we did”.

Even if you engage the early adopters, you’ve still got to cross the chasm – turning your back on those early adopters to capture the majority of the market.  That is tough when you are trying to deal with customers who have paid for their offering and may not be happy with a sudden change in direction.  (Then your engineers decide that if you are going to change the product that much they will need to redesign it from scratch.)

Meanwhile, a smart “fast-follower” (sometimes called “fast second”), can watch, learn, refine their product and then pounce once the market is understood.  In Lieberman’s study he found that “in the early years” of the dot-com boom “brick and mortar hybrids had substantially higher revenue than the average firm in their market space”.  Maybe that was because, unlike all the new entrants, they actually understood their customers.  Radical idea!

What should we take away from this in considering our product strategy?

1. Be aware that you are balancing the benefits of being the first to market with the investment in developing a complete product for a market you do not yet understand.
2. Jump too early and you’ll be tied down by your early customers and product and find it difficult to shift; the “winner’s curse”.
3. Jump too late and the early entrants will have considerable (perhaps unassailable) lead in product development and market mindshare.
4. Be nimble.  Be ready to recognise that the market is shifting and adapt your offering to it.
5. Don’t give up if you are beaten to the market.  Watch what happens very closely, talk to the early adopters, try to identify the early majority, and build a deep understanding of what is going on.
6. Don’t get stuck servicing your early adopters, they can sentence your product to a life in a niche.
7. If your product needs “network effects” make sure you have a strategy and the resources to be the first to get to the tipping point; then the market is yours.
8. Ensure any “switching costs” offer benefits for the customer which outweigh the costs to them.
9. Patents?  Innovate, innovate, innovate…
10. Be lucky!

Related posts:

What to do in a Recession? The Tyranny of the Served Market Be Careful What You Wish For… Hierarchy of Product Features Going ‘Off-Strategy’ Projecting simplicity