The Startup Genome Report investigated over 3,200 high growth tech startups. The findings revealed that over 90% fail, in most cases due to self-destruction rather than competition. But perhaps the greatest insight was that 74% of startups fail due to premature scaling. Avoiding premature scaling has become one the most pressing imperatives for founders, yet many don't start managing this risk until it’s too late. What is premature scaling and how do we avoid it?
Understanding the startup lifecycle is the foundation for assessing the state of progress of any new business. Any startup that does not have a clear appreciation of the key stages of the life cycle is at high risk of premature scaling. These 6 stages, taken from the Startup Genome Report and known as the Marmer stages, are:
In our earlier article on funding milestones, we mapped these first 4 stages onto a simple progress chart, the Customer Development Process, describing the gates an early stage business must pass through to transition from one stage to the next. For each gate we must set specific thresholds to test whether or not we should graduate to the subsequent stage.
The first three stages are where premature scaling will most likely occur: in such cases investments in subsequent stages are made before the current stage has been completed. We get ahead of ourselves, burning cash on what seems like the right things, but in the wrong order.
For example, in the Discovery stage the focus must be on searching for the Problem/Solution fit. Is there a big problem worth solving? In addition to really nailing down the target market size, the ultimate test is positive early adopter reaction to a minimum viable product (MVP). Classic mistakes here for example are (i) building out a stellar MVP without being confident that the opportunity (market size & timing) is even worth it, and (ii) investing too heavily in what we believe will be the killer product without sufficient (or any!) early adopter feedback.
In the Validation stage, we are seeking to confirm Product/Market fit and test the Business Model assumptions, such as price. Classic mistakes here are (i) scaling the commercial team before the product is ready, and (ii) investing in production capacity that exceeds early scaling volume requirements.
In the Efficiency stage (also known as the Customer Creation stage), we must be confident the Go to Market strategy is working, i.e. a repeatable sales model has been found enabling initial scaling past the early adopters into the early majority. Classic mistakes here are ramping market development activities in new sectors or territories before (i) the initial beachhead market is secured, and (ii) all the critical KPIs (e.g. CAC, retention, COGs) are tracking to expectations.
Different types of startup
To determine the specific thresholds that our business should pass at each stage, we must first understand the type of business we are creating. What I mean here is the type of Product and type of Channel though which the product (or service) will be sold.
For example, will this be a company that builds physical products and sells them through a physical channel? Or a company that will build virtual products (e.g. software) sold through a virtual channel (web/mobile)? The timescales and costs will vary considerably between these two extremes. This will have a huge bearing on capital intensity, how much investment should be raised at each stage, and how investors will assess the risk/opportunity profile of the business. Hence the need to set different thresholds for stage transition.
Steve Blank provides an excellent breakdown of the four key types of business based on this Product v Channel model. These can be shown as follows:
In the past decade the number of companies developing virtual products for virtual channels has exploded. For example, according to Statistica, public cloud SaaS was a $8B market in 2009 and this year will top $141B. VCs have embraced this new paradigm with unrelenting enthusiasm much to the cost of more ‘traditional’ businesses building physical products, where the cost (and risk) to scale can be orders of magnitude greater.
By applying the quadrant model, we can set specific thresholds that our type of business must pass in order to progress forwards. These thresholds should ideally correlate closely to the investment criteria used by VCs and other investors, and these criteria can often be found on their websites. Many are explicit and will describe what they are looking for at Seed stage, Series A, Series B and so on. This requires some legwork but it is essential to take these criteria into consideration when setting objectives for each stage of company development, as these stages can provide the most natural funding points.
Assessing premature scaling
Avoiding premature scaling is part science and part art.
The science is to ensure that the cash burnt on delivering each phase of company development is substantially allocated to meeting the exit criteria of each stage.
Therefore, keeping track of how cash is being deployed in high burn areas like product development, customer acquisition, and team expansion - during each stage - is critical.
In the Startup Genome study, these high burn areas are referred to as ‘dimensions’. Inconsistent startups - those most prone to premature scaling - have one or more of these dimensions far ahead or far behind the Customer dimension. The key dimensions and examples of their associated risks are shown in the table below.
‘Platform’ businesses are highly prone to premature scaling
Whilst any type of early stage business can fall into the trap of early scaling, platform businesses are particularly prone. I am referring to businesses where the technology can address multiple use cases.
The risk here is that whilst in search of the ‘killer app’ the company invests time and money in creating solutions for multiple applications yet fails to find a solid Product/Market fit for any one significant business problem.
So, whilst the company is essentially still in the Discovery phase its behaviour in terms of spend is that of a Growth stage business that is developing new application areas beyond its beachhead market.
Investors are hugely wary of early stage platform businesses for this very reason. This is highlighted in a recent article by Philip Lay for Octopus Ventures. The advice for platform business here is “In strategic terms, your first job is to identify a compelling use case — preferably in the form of a broken business process that causes your target customers unacceptable levels of grief — and then to make sure the problem is fixed with as complete and repeatable a solution as possible.”
Hardware businesses developing embedded or component technologies should particularly heed this advice. Over the past 10 years we have seen numerous electronic component businesses suffer from this type of premature scaling. This is in part caused by not raising sufficient Seed capital to see the business through Discovery and Validation stages to confirm the beachhead application and prepare for initial scaling (Efficiency). As the money runs thin and the Product/Market fit in the initial sector remains elusive, the business tries in vain to target new applications…. before realising it’s simply too late.
74% of high growth startups fail due to premature scaling.
Understanding the startup lifecycle stages and the criteria for passing through these stages is the foundation for avoiding premature scaling.
Setting clear thresholds for stage transitions is essential and cash burn should be allocated heavily to meeting these criteria.
Aligning these stages with typical investor funding points for your sector and business model requires legwork to research but pays huge dividends.
As a result, startups that scale properly grow about 20x faster than startups that scale prematurely.
And remember, the seeds of destruction are often sown early in the Discovery stage. It is said that successful startups succeed because they are good ‘searchers’ and failed startups achieve failure by efficiently executing the irrelevant.
What kind of startup are you?
About the author: John Hall is CEO and co-founder of Duet Partners. His 30-year tech career began with major US semiconductor and software companies, and was based in the Valley during the late '90's. Before Duet he was CEO of a VC-backed consumer electronics company, sold in 2009 following several rounds of capital raising. In the past 10 years since starting Duet he has advised dozens of founders on the startup to scaleup journey and is a retained Board advisor to a number of UK technology companies.