The Elusive Next Big Thing: Part 1

Inevitably, whenever I spend more than a few minutes with a Venture Capitalist, entrepreneur wanna-be, or the insurance salesman down the street, I get asked the question “what’s the next big thing?”    Although unstated, I think this question really expands to “what’s the next big thing in information technology that will create opportunities for start-up companies to emulate the success of Cisco?”

Unfortunately, my well researched and well considered answer is “I dunno”. 

First, let’s define the necessary backdrop for start-up success. 

Specifically, let’s narrow this down to start-ups aspiring to build systems or software products that leave their facility ready for end-user deployment (even though the products may flow through a sales/distribution-oriented channel before arriving to the end user).  The end users in this case can be enterprise or service providers.  The success formula for “arms dealers,” which include semiconductor vendors, optical components vendors, passive component vendors, and software embedded protocol/stack/feature vendors; is very different than for systems and packaged software vendors.  We’ll leave the discussion about arms dealers to a different day.   The dynamics for various forms of service providers are also different and therefore not addressed here.

Furthermore, for different reasons, let’s limit this discussion to vendors in the network and application delivery infrastructure space, simply because that’s all I know about.  I don’t have much experience with enterprise software, for instance.  I also don’t know too much about consumer markets.  The dynamics may be quite different in those spaces.  I simply don’t know.  But my intuition tells me that the equation for start-up success isn’t all that different across market types if the start-ups share the goal of introducing deployment-ready products into their chosen market segment.   I’ll let those of you more experienced in those spaces decide what rules outlined here apply.

What I do know is that for a systems start-up to succeed with targeted enterprise or service provider customers, they need the following factors working in their favor:

1. A start-up needs to introduce a product into the market at a very specific phase of a very specific market growth curve.  That is, the start-up needs to introduce their initial product at an early enough stage of the market growth curve to gain significant market share and become a recognized leader in the space, but late enough so that market expansion occurs soon thereafter.

At the time of market entry, there must be enough willing and pre-conditioned early adopters to deploy the product, to help the  start-up “harden” the product, and to generate enough first shipping year revenue to allow the start-up to stop burning investors’ cash and/or to raise more investors’ cash. 

A few years ago, when many target customers acted like early adopters, it was almost impossible to bring a product to market too early.  Time-to-market was everything.  Nowadays, when things are more normal or even tweaked towards customer conservatism, it’s very possible to miss-time the market early.  When this happens, start-ups usually find themselves in a very awkward position.  They are burning increasing amounts of cash on customer support and marketing, but not generating enough business to cover those additional costs.  As a result, what usually happens is that R&D effort is cut back and slowed down.  The problem is that technology rusts.  If, in the future, the market does start to expand, the start-up might find that the technology introduced a year or two ago doesn’t meet the needs of the day, or not as well as a more recently introduced product from a competitor.  Slow growth markets are death to early entry start-ups.  Ironically, in a slow-growth market, the last one in before market expansion often wins.  Worse yet for start-ups, the last one in may not be another start-up but an established vendor with an existing channel to leverage.

2. A start-up requires, in the early year(s) of market penetration, customers that see a high value in their product and are willing to pay a premium (read high gross margin) for their solutions.  Financial necessities prevent systems start-ups from participating in markets that are already or are rapidly being commoditized.  You can’t “get it back on volume” when you have all the costs and time constraints associated with building a channel in front of you.

3. A start-up solution needs to be deployable in small chunks and still deliver value to the customer.   If the value is great enough, the customer will deploy more and more of it and others will do the same.  This kind of grassroots adoption and wildfire expansion is typical of almost all successful new technologies, new product categories, and specific products in the networking space that I am aware of.  It’s death to a start-up if the customer only receives value if they deploy the solution on a large scale, the worst case being end-to-end.   A great example of this is ATM, which never took off in the enterprise space because it didn’t really deliver any value when deployed on a small scale relative to the cost and churn.   But ATM found a home as a point application in the carrier space as a backbone to scale Frame Relay networks, and then grew from there.

4. A start-up needs to bring to the market a sufficiently different and (most likely) disruptive idea, to justify their existence.   Customers will simply not buy the same-old-same-old but at a lower cost from a start-up.   Lowering the cost of an existing well-understood widget is a non-starter.  So is simply integrating several well-known functions into a single box or steaming pile of software. But radically reducing costs by eliminating the need for a bunch of widgets previously considered to be essential can be a very interesting value proposition, if this represents the result of a new way of thinking about application or network architecture.

5. Unfortunately, the argument above needs to be won largely on a CAPEX analysis only.  TCO and OPEX arguments can augment the CAPEX argument, but can’t be the critical component of the economic value proposition.   Start-ups can never play the “easier to manage” card (unless their product is a management application, I guess, but don’t get me started on that!) because the process of evaluating a start-up and its products, testing it, integrating it in to the infrastructure, etc. are in itself a huge management burden.  Also, the old stuff doesn’t go away. So the start-up product just represents yet another thing to manage and another complication to overall system management. 

The trick is to be an overlay that adds a whole lot of value and sets the direction for the future migration of the infrastructure without disrupting the function and management of the installed base.  It’s a pretty high bar to hurdle.  So changing the definition of the game in some fundamental way is crucial.  In fact, it has to be changed enough so the existing vendors fight against it because of the impact such a disruption could have on them.  If their response is “good feature idea, we’ll have it in our products within 6 months” and the customers view that as a credible promise, then the start-up is dead.   Instead, the response of the existing vendors needs to be “that’s crazy,” or “that’s stupid.”  Kind of like the SNA vendor and PBX vendors back in the days when start-ups suggested that the future of networking was TCP/IP and Ethernet. 

6. A start-up needs (a little) competition. When changing the definition of the game, one needs a loud voice.  The natural reaction of target customers will be caution and cynicism.  This is true of analysts and the press as well.  Unless they hear the same story from multiple sources, over and over again, the ground-breaking concept will get lost in the noise of all of the other information promulgated in our industry.  The best way to ensure that several entities are taking the same message to the world is to have several entities with similar self-interests spending energy on that.  That generally means some competitors.  Although it is out of any start-ups control, the ideal number of early entrants in a new market or product space seems to be about three to five.   That means three to five vendors funded to do about the same thing in the same six-month time span.  Over time, as the story gets out, this space will get “hot” and many more aspirants will be launched into it by the VC community.  Depending on how the market timing works out (see point #1 above), these newly launched companies will either be “dead companies walking” because they missed the market introduction window, or they will represent the next group of viable aspirants if the market develops slowly and the initial group fades away (possibly to be replaced by yet another set six months later if the market continues to be slow to develop; assuming the space hasn’t yet gotten branded as “evil” by the VCs).

7. A start-up needs great investors (patient, helpful), a great team (experienced, flexible), and most of all, lots and lots of luck.

So, given these criteria for the Next Big Thing, at least from a systems vendor start-up point of view, what are the present candidates and how do they match up to these criteria?  I’ll examine that in next month’s Musing, starting with a little historical analysis of the last Next Big Things and what factors determined their success or failure.

(volume 6, number 9)