New product / new service introductions in companies that have an established portfolio is another one of those “hard” things. No one wants to kill the goose that’s laying the golden eggs. No one wants to be responsible for moving resources off of projects where there’s a clear return on investment – or more importantly a perceived return on investment. It’s tantalizingly simple to say “I have a customer that wants feature A so let’s add in feature A – nothing more, nothing less”. For you Office Space fans, it’s the Tom Smykowski school of product management – “…you take the specifications from the customer and bring them down to the software engineers…” It’s a viable shortcut and an expedient way to address the want – not necessarily the need. This approach is fine when you’re in a high growth market with an established position – the cash cow. Ride that growth and don’t make any waves of your own. It’s prudent and logical, in these cash cow situations, where you’re focusing on economies of scale to improve your bottom line primarily by what you can control – cost. But what if you’re not managing a cash cow or what if your cash cow is becoming a dog? Woof followed by oof. Where and when do you find the next growth curve? Or do you, like Snow White, hope that someday your growth will come?
Pivot is not a four letter word. It’s a recognition that your current approach isn’t providing the growth that you need for a viable business, and that it’s now time to switch it up. So, how do you pivot? How do you get that new growth curve going? This is where the Fool’s ROI does the most egregious damage. What do I mean by Fool’s ROI? Well, first, let’s define ROI. ROI = (Net Profit / Cost of Investment) x 100%. Simple enough, but stay with me as I get into a wee bit of algebra. If one looks at ROI through the lens of comparing variable returns based on a constant investment versus fixed returns based on variable investments, it looks like this:
The yellow line shows the linear improvement in ROI for a fixed level of investment. That return is based on market acceptance of your product, which means you’re taking a dependency on your customers, so that’s out of your direct control. The red ROI line is asymptotic and you’ll see the value approaches infinity as the cost approaches zero. And you’re in direct control of that cost, so break out the champagne as you’ve cracked the code, bud – near infinite returns from a NO investment decision! That is the Fool’s ROI – do nothing and automagically you get an infinite ROI. It’s a formula, presented as sage wisdom and sound business acumen, but in fact it is really a mathematical cover-up for the crime of cognitive biases toward inaction. The primary cognitive biases that I’ve encountered are:
- Status Quo Bias – “Stems from our tendency to avoid losses and regret at all costs. Since deviating from the norm is risky, we often stick to inaction.” Even when the results of inaction can be measured as detrimental to the business.
- Omission Bias – “Occurs because we overgeneralize the belief that actions cause more harm than omissions. Additionally, when we act and cause negative outcomes, we view that as a greater loss than when we fail to act and cause negative outcomes”.
- Loss Aversion – “Loss aversion is a cognitive bias that explains why individuals feel the pain of loss twice as intensively as the equivalent pleasure of gain. As a result of this, individuals tend to try to avoid losses in whatever way possible”.
In lieu of rewiring our personalities to avoid these biases, how do we recognize and work through these biases when making product, solution and go-to-market decisions?
The first step to recovery is admission. Have you really done the analysis that shows that inaction will result in a worse position for the company or are you providing excuses for yourself via the cover of a “guaranteed return”? Remember, as I stated in the previous blog, there ain’t no guarantees.
Once you recognize your bias, you can have an honest discussion with yourself. Note you don’t need to stand in front of a mirror for this, but you do need the time to reflect. Are you a fast follower (good money to be made here, no shame) or are you an innovator (high reward but has risk)? Be really honest here. Everyone says they want to be a market leader, but being a fast follower can be very lucrative as you don’t incur the same investment risks with having to make/lead the market. If you can’t be a fast follower because your market is a question mark, perhaps your market is entering the trough of disillusionment; you’re going to need to innovate to differentiate. Innovation, not only in the engineering sense of deploying a new technology in your chip, but also based on your whole product / solution, and even the ability to differentiate via business models.
Now is the moment of truth. You have a hypothesis that requires investment to realize those anticipated returns. Oh, there’s that modifier – anticipated – not actual. This hypothesis shows that there may be a market, there may be a persona, there may be a use case and this may be your total addressable market (TAM), may be your serviceable addressable market (SAM), and most importantly may be your serviceable obtainable market (SOM). A lot of maybes. This uncertainty, due to lack of a stack of purchase orders in your pocket equaling your SOM value, when mixed with omission bias, leads to the facile decision of NO.
So how do you push through the NO? Not necessarily to the YES but to the non-binary MAYBE which leads to that data driven YES or NO. I’ve found that the lean startup approach is a great technique for pushing through the NO to a more nuanced position. For software companies, especially SaaS companies, this section isn’t for you. This is in your DNA. I’m talking to you semiconductor companies! 7nm and smaller geometry chips can cost 100s of millions of dollars. Decisions are obviously costly. Chip design for decades has developed a set of tools and methodologies for mitigating risk. The chip development workflow is a classic stage gate waterfall with numerous checks and balances. There’s no way to do “agile” with a tapeout…at least yet. Unfortunately, that methodology can permeate company culture resulting in tapeouts being treated the same way as social marketing decisions around tweets. This results in every decision being treated as existential and going through things like business planning processes for even immaterial sums of resources resulting in higher personnel costs incurred with the decision making than if you cut the actual check needed for the execution!
This is the lean startup methodology:
Build, Measure, Learn then rinse and repeat. You’re building a Minimum Viable Product (MVP). The MVP isn’t the final end product. It’s the offering that happens at the bottom of the funnel that is for validated learning. What do I mean by funnel? Let’s go with the AIDA model of a funnel as an example:
To get meaningful market feedback that allows you to get to the ultimate offering that transacts in volume, you have to interact with customers around a representation of the final offering – the MVP. You will not be able to get away with avoiding engineering complexity by staying higher up in this model in the “Land of Slideware”. So, if you keep costs down by updating your web pages and focusing on search optimization, you may achieve Awareness, but Awareness only means that the customer knows you exist. That’s one-way communication, so you’ve got to get through Interest and Desire. Product briefs, white papers, customer presentations, demonstration videos and competitive analyses are tools for moving customers through Interest and Desire but Action is where the interAction occurs.
InterAction at this point is not the ultimate financial TransAction but all about Validated Learning via the MVP. According to Eric Ries, “The goal of an MVP is to test fundamental business hypotheses (or leap-of-faith assumptions) and to help entrepreneurs begin the learning process as quickly as possible”. First step is you need to Build and this is where you can get through the NO by something nuanced.
Here are a few examples:
| Hypothetical Ultimate End Product | What you build for the MVP |
| An ASIC for AI Inference Acceleration | An FPGA for AI Inference Acceleration |
| A bespoke chip down appliance based on your ASIC | Leverage your existing PCIe board and use a 1U server |
| Needs Windows OS | Assuming you only support Linux, stick with Ubuntu. Windows can come if there’s enough interest. |
| Top market share ISV as part of your solution (who as of now won’t return your calls) | An ISV that’s as hungry as you are and willing to invest in a joint offering. Don’t worry about market share – worry about what you’re delivering for the user experience. |
Building your MVP allows you to measure market acceptance via customer feedback. Leading key performance indicators (KPIs) are crucial here – NOT just design win opportunity dollars! You’re looking for validated learning as measured against KPIs – too early to ring the bell on those big deals yet. That time will come! By the way, this is something that SaaS companies do really well – customer acquisition via freemium models for example. Acquire customers today…monetize tomorrow. Once you’ve Measured, you now have the data to do your gap analysis. And from that analysis you Learn. The ultimate Learning gets you to where you know that by closing these final gaps you can transact. Or you may find out that the market isn’t interested in what you’ve got at all – giving you the data that you need to pivot. Now you’re making decisions based on data and not bias. And you’ll build a little customer intimacy along the way.
So let’s not be fooled by the Fool’s ROI. Inaction may feel safe, even smart, when couched in the language of return-on-investment, but that safety is often an illusion, propped up by biases we don’t even realize we’re carrying. The real risk isn’t in the doing – it’s in the waiting. When you build, measure, and learn, you’re not just gathering data, you’re reclaiming control. In a world full of unknowns, progress belongs to those willing to move, however small the step. So build that MVP, get it in front of real users, and let the data tell you whether to double down or pivot. Either way, you’ll be moving forward, and that’s a return worth investing in.
Philip Lewer
Managing Partner
Soundview Consulting 