Rethinking R&D Strategy
Kobus Cilliers | On 05, Apr 2020
â€œResearch is what I’m doing when I don’t know what I’m doing.â€
Wernher von Braun
Look at most R&D strategies and one of the most likely things you will see is how incremental and focused on obvious directions they are. Make things faster, lighter, thinner, last longer seems to be the expected norm. All very sensible things to focus on (providing, of course, that your customer actually still wants more of what you gave them last time around), but sadly not a great way of designing an R&D strategy. Because they tend to be uni-directional and thus end up hitting brick walls. The material gets thinner but less durable. Or less dense and more expensive. R&D strategies generally focus on components and their attributes (â€˜thingsâ€™) rather than the interactions between them. The â€˜betweensâ€™ being the contradictions. Which, as we know from all of the TRIZ/SI research is one of only two ways to innovate. The other one being adding a new function or outcome to an existing offering. Maybe, we hypothesise here, if these really are the only two ways to innovate, surely they also show us how to structure our R&D programmes?
This shift from â€˜the thingâ€™ to the interaction between the things has subtle but â€“ we propose â€“ quite profound implications for R&D strategy. But, as is almost inevitably the case, nothing is ever quite as simple as it sounds. A lot depends on the capability of the R&D team and the organisation around them. They might all be better off focusing on the â€˜betweensâ€™, but the types of â€˜betweenâ€™ they should focus on will be different according to the available capabilities. The main idea of this article is to try and map out a few R&D strategy design heuristics as they relate to the different Levels of capability as defined in our Innovation Capability Maturity Model:
Weâ€™ll examine each of the Levels in turn, and try to best define where an R&D strategy might come from and what it might look like. Letâ€™s start with the simplest:
ICMM Level 1
Level 1 is the simplest because the only real imperative are for an R&D team to demonstrate some tangible success and create a â€˜sense of progressâ€™. What the success actually looks like is largely irrelevant since the primary job is building trust between the team and the rest of the business thatâ€™s paying for them to play in their sandpit. What weâ€™ve found ourselves doing with several of our ICMM Level 1 clients is finding problems and ranking them in terms of â€˜deliverabilityâ€™. Far better at Level 1 to finish something small than fail at something bigâ€¦ even if that â€˜failureâ€™ might be a planned learning step along a longer journey, the rest of the organisation wonâ€™t understand the concept.
Finding problems in a Level 1 organisation is easy because theyâ€™re everywhere. Where there is a desire to bring some structure to the search, and help plan for a more structured future, the most effective means of creating an R&D strategy is to do it on a silo-by-silo basis. Which, given that Level 1 companies are almost inevitably â€˜Operational Excellence driven optimizersâ€™, will mean lots of lots of minutely siloâ€™ed individuals. Innovating across silos is hard. Most innovation opportunities, therefore, are going to come from staying within your own silo. Being systematic within your own silo then typically means constructing a Function and Attribute Analysis (FAA) model of what you control. Probably a â€˜sub-systemâ€™. Construct the FAA model, and â€˜R&D strategyâ€™ effectively means looking at all the negative relationships, working out an â€˜ease-of-fixâ€™ ranking protocol, and working through the resulting list.
ICMM Level 2
By the time Level 2 has been attained, sufficient of the silo walls have tumbled to consider constructing system-level FAA models. And to be a bit more objective about ranking which negative relationships in the model deliver the most impact to the system as a whole. There is still likely to have to be a certain degree of pragmatism applied in terms of establishing whoâ€™s on the innovation journey and who isnâ€™t, and defining projects according to the resources at hand (â€˜critical mass at the critical pointâ€™), but there ought to have been a clear jump beyond the â€˜whatâ€™s easiestâ€™ selection criterion applied at Level 1.
Particularly brave Level 2 organisations might also consider building tangible/intangible customer outcome maps into their portfolio of R&D project definition tools, thus opening up the possibility of establishing R&D projects focusing on delivering new (or more likely, â€˜hiddenâ€™) customer outcome needs. In several industries, there is a lot of low-hanging fruit right now in terms of unmet â€˜intangibleâ€™ needs. Thereâ€™s probably enough, in fact, to consider making it a â€˜standardâ€™ ICMM Level 2 R&D strategy project opportunity finding tool.
ICMM Level 3
A Level 3 organisation is characterized in part by the disappearance of internal silo walls, including the one between Operational Excellence people and the R&D team. At least in terms of recognizing that the two need and benefit from each otherâ€™s existence. The only silo wall that now hampers the definition of a meaningful R&D strategy is the one between people on the inside of the organisation, and all those outside. From a very practical perspective, this means the â€˜domain expertâ€™ still rules the roost. Which in turn means that R&D strategy is most effectively built around advancing from todayâ€™s knowns. A suite of FAA models for each product, service or part of the enterprise will offer up a catalogue of R&D strategy components, but these can now be complemented with overarching hierarchical models that R&D strategists can use in conjunction with a â€˜Trimmingâ€™ perspective. Components at the bottom of the hierarchy become candidates for an R&D project to migrate their function into a higher level sub-system.
Also relevant to introduce into the R&D strategy definition process at Level 3 is the IFR Attribute map:
What this tool brings to the story is a comprehensive means of identifying â€˜allâ€™ of the contradictions and unmet needs within a system. R&D strategy then basically distils down to tackling each of them in turn. The IFR Attribute Map also allows ICMM Level 3 R&D strategists to more objectively prioritise the various contradictions that get revealed in terms of their relative benefit to customers: CustomerA-versus-CustomerB contradictions being top of the impact list, followed by AttributeC-versus-AttributeD conflicts, followed by Customer-versus-Provider contradictions.
ICMM Level 4
If the dangerous concept of Ideal Final Result needs to be covertly snuck into the discussion in a Level 3 organisation â€“ i.e. by focusing it on individual attributes of a solution rather than at the macro level â€“ a Level 4 organisation is mature enough to know that it is the overall Customer Ideal Final Result that the R&D strategy needs to stem from. Which in practical terms means that the strategy becomes all about filling in all the details of a â€˜Function Coneâ€™:
The reason this everything-converges worldview is something a Level 4 organisation can socialize whereas Level 3 cannot is because a Level 4 organisation has successfully solved the â€˜domain expertâ€™ problem. Domain expertise in a Level 4 organisation is about Function-based domains rather than solution based domains. A Level 3 lawnmower company will find it very difficult to build a meaningful R&D strategy based on a â€˜grass stops growingâ€™ end point because all the employees tasked with making better engines, blades and grass-cutting holding capacity will see it as an attack on their expertise. People in a Level 4 company, on the other hand, will recognize that grass that stops growing is inevitable and â€˜better for us to do it than somebody elseâ€™. R&D strategy in a Level 4 organisation is also likely to make use of TRIZ Function Databases for the first time, and building at least a part of their R&D strategy around exploration of other means of delivering the required functions.
ICMM Level 5
Not that there are many exemplars of Level 5 organisations around to go and talk R&D strategy with, but one thing that characterizes them over a Level 4 organisation is that while the Level 4 R&D team will allow themselves the luxury of drawing one Function Cone, the Level 5 will very likely have built their R&D strategy around a constellation of different Function Cones. The general idea here being that, once they calibrate each of those Cones in terms of their pulse rate and â€˜ease of disruptionâ€™ theyâ€™ve pretty much built themselves the ultimate map of the whole R&D world. And done it in such a way that they can focus available resources to consistently deliver the best â€˜bang per buckâ€™.
Where â€“ to probably hammer the initial point home one too many times â€“ that bang-for-buck calculation ultimately comes back to finding all the â€˜betweensâ€™ – the Contradictions and the unmet customer needs – and going after them in a structured, repeatable manner. A bit like the mouse trying to reach the cheese at the end of the maze. Only, in the case of Level 5, where the mouse gets to decide which mazes to play in, and which to avoid.