Feature

posted 25 Sep 2002 in Volume 6 Issue 2

Knowledge management for new ventures

KM has traditionally been concerned with efficiency in large organisations, but how relevant is it to small start-up companies focused on growth? Sam Marshall examines the unique knowledge risks and opportunities that set new ventures apart from established big business, and how KM practices can be adapted to cope with these fast-changing environments.

Imagine working in an organisation where you knew exactly who knew what, and what everyone knew; where all operations were transparent; where changes could be brought about in a matter of weeks; and, where everyone shared knowledge openly because you were all working towards a single goal. This can be the reality of knowledge management in a small start-up company like a new venture. Now imagine being responsible for managing knowledge in a company where 50 per cent of your staff is new to the organisation; where you could lose several entire departments overnight due to illness; where most of your operations have never been done before, so processes are invented on the spot and under pressure; and, where some of your key functions don’t belong to you and could suddenly start working for a rival. This, too, can be the reality of KM in a new venture.

This article focuses mainly on corporate venturing. Many large organisations look to spin-offs, or new ventures (NVs), as a way to keep a stake in a business but unshackle it from the limitations of its parent. They pick a small team of capable entrepreneurs and give them plenty of space.

Knowledge is essential to NVs, but to apply KM thinking you have to re-scale the toolkit appropriately. In a recent project by KM practitioners in Unilever, we found that, in general, NVs have some naturally effective KM practices, lead by entrepreneurs that intuitively know they have to learn to survive. But the crunch can come as the companies grow, so we’ve been experimenting with applying some pre-emptive KM thinking to ease the growing pains.

What makes NVs different?

NVs face a special set of problems that come from operating in a new domain with limited resources. Figure 1 illustrates these problems from a knowledge perspective. A new operation is not just about doing, but also about learning to do. It needs to build knowledge about creating its products or services, as well as to establish normal operations, just as a novice carpenter has to simultaneously run a business and learn how to create new designs.

In The Alchemy of Growth[1] Baghai, Coley & White describe managing over ‘three horizons’. Horizon one is about making money from existing business; horizon two is about emerging business; and, horizon three is about future options for growth. The key, though, is in managing all three at once, so that there’s a steady renewal of business. It’s tempting for a NV to see horizon one as the most pressing issue, focusing on operational issues and the bottom line, but this could mean the company lives only as long as its first product. In reality, the launch phase of an NV is the first transition of an idea from horizon three to one. Each transition into a different horizon means the management has to learn how to operate in a new way. Some NVs never learn how to change mode to suit the horizon they are facing.

In this sense, larger, established businesses have it easy: they already own the knowledge for normal operations (horizon one) and so can focus on refining them. They probably also have dedicated development staff for horizon three, and the unknowns are far fewer. By contrast, setting up a new venture involves stepping out into a whole world of unknown quantities.

When it comes to human resources, NVs have to handle massive proportional growth (staff numbers can easily double every year for the first few years). This rapid growth can be the undoing of some, because it’s much easier to hire someone than it is to assimilate them effectively so that they can apply their knowledge. Often employees also cover many roles; one person leaving could represent three whole departments worth of knowledge walking out the door. On top of this, we’ve noticed a tendency in NV management to overestimate the competence in their team members to do all these tasks equally well.

Figure 1 – new ventures versus established businesses from a knowledge perspective

How can KM Help?

Our thinking on KM for NVs has been formed through a number of contacts both inside and outside Unilever. For illustration, however, we’ll focus on one case, which was a move into the confectionery market by Hindustan Lever Ltd, Unilever’s Indian subsidiary. The confectionery NV has been running since late 2000, and reached the number two or three position in the markets it entered within five months of launch. The business began with just a leader, soon growing to a team of four. Supplementing this was a large number of partnerships, both internal (for instance with market research managers in ice-cream and toothpaste divisions who also target the child consumer) and external (for example product development, advertising and manufacturing). Additionally, the leaders of several NVs within the company formed a community, which allowed regular opportunities for them to tap into each other’s tacit knowledge.

The challenges faced during initiation were substantial. Which, of a number of possible products, should they launch first? How were they to compete with established players? After the first launch, how would they scale-up to new regions and a bigger team? How would they come up with not just an initial winning product but also an ongoing succession of them to keep their fickle market (children) interested?

In helping an NV at this stage, KM has three main tasks:

Manage knowledge flows over the extended team;
Manage knowledge flows over time;
Manage knowledge flows over the three horizons.

Managing knowledge flows over the extended team

Mapping knowledge

NV teams may look small, but they extend widely through the people that get involved. This ‘invisible’ team may not even be officially part of the NV, but it helps it to operate through a network of favours to the team members. In established businesses, defined workflow, co-location and management teams would create lines of communication. As these rarely exist in an NV, we favour workshops for KM interventions because they help extended team members feel more integrated into the overall task. In effect, the structure becomes more like a network and less like a hub-and-spoke design formed around the core team. For example, knowledge mapping workshops can help NVs on a number of levels: they help to visualise what is known, who knows it and when the team might exploit it; extended members also see what the others do, and what’s required or expected of them. Mapping also shows what is not known (which may not seem discussable when somebody is providing a favour), and so stimulates appropriate planning. In one workshop we ran we asked the extended team to say ‘who knows about...’ and listed out some key topics. This revealed that the knowledge for a whole area of operations rested with just two people, who in turn were considerably alarmed to see what they were expected to have answers for.

Extended team challenges

As the NV grows, knowing how everyone’s knowledge fits in becomes even more important, particularly when formal agreements for services need to be set up. Tasks may be done once or twice as favours, but eventually they need to be properly sourced and suppliers briefed on exactly what is required. Similarly, new recruits to the core team can perform much more quickly if they can see who has provided what expertise in the past.

A downside of extended teams is that key knowledge can be held outside the NV, and thus outside its control. Using an intellectual asset perspective can help identify the capabilities that give the NV a competitive advantage, and those that can be safely outsourced. The confectionery NV, for example, leaves manufacturing to third parties because it is product design and marketing that really matters. NVs always need to use third parties to get started. What helps incubate NVs internally is that capabilities developed by partners in other parts of the parent company will not subsequently be used to help competitors.

Another aspect of reliance on favours is that it implies that a predictor of the NVs success will be the networking abilities of the team members. Mapping personal networks could be used as a screening tool by investors and a recruitment consideration by the team. It’s common for venture capitalists to assess the networks of entrepreneurial teams, but corporate venture units tend to ignore this and let the person who had the idea lead the team.

Managing knowledge flows over time

The concept of versioning

Managing flows over time doesn’t mean just preserving knowledge; it means ensuring that past knowledge can be re-challenged, updated and refined. One way to do this is through ‘versioning’. This term is taken from a concept in iterative methodologies (as in software engineering) where the evolution of a development is managed by taking periodic snapshots (eg, versions 0.1, 0.2, 0.3, 1.0, etc). For a new venture, this means tracking the business’s understanding as it grows so that there is always a current view of ‘what we do’, almost like an ongoing autobiography. The essence is to build in a ritual of updating the content to keep it in step with what actually happens.

To encourage the confectionery NV to make use of versioning, we attempted to introduce a versioning tool. This was a hierarchical browser, similar to the file explorer in windows. The hierarchy, however, represented the tasks and sub-tasks required for running the business. For each task or key decision, users could add resources such as presentation files, comments about usefulness, more general learnings and the roles people had performed in fulfilling a task. They could also add the kind of ad hoc tips that would otherwise get lost because they don’t warrant a full document. This was coupled with a process of regular debriefs to elicit further content. The tool offered many other tricks such as the ability to ‘reverse search’ an entry and all the activities that a team member has been involved in, which was also useful as a dynamically evolving job description should new members take on some of these roles.

Why is versioning so necessary? Decisions made in the past are the product of an NV’s best understanding at that point. If, say, six months down the line the NV has to explain a decision, it can be hard to recollect exactly what lead to the conclusions that were drawn. This introduces the risk that when the NV repeats an activity (for instance the launch of a second-generation product), earlier decisions are not challenged, even though more recent knowledge would indicate that a different decision would be more appropriate. All too soon things are done one way ‘because that’s what we’ve always done’. A participant in an early workshop that we conducted said it was a revelation to her that most of the team’s knowledge was “not at all well established, but in fact anecdotal, personal belief and undocumented”. It’s hard to challenge decisions based on knowledge like that. Moreover, humans tend to re-sequence and re-structure events to post-rationalise what they do. This is because they rarely track what they know at any one time, and are not necessarily aware of it later on (instead this knowledge remains tacit, often explained as gut feeling). Semantic memory (what is) operates quite differently to episodic memory (events). We all know what the capital of France is, but we can’t necessarily say when we learnt it. So people usually cannot say exactly what semantic knowledge they had when a particular decision was made.

Making versioning work

There should be few surprises here for the KM veteran: to get versioning to work you need an easy-to-use tool that’s embedded in the organisation’s processes, adds value and is recognised as being important.

As always, the big barrier is time, due to the constant pressure to perform that NVs are faced with. One solution is to assign an explicit knowledge assistant role, someone charged with maintaining the versioning tool. Everyone in the team is still responsible for the quality of what’s captured; the assistant merely creates the stimulus (for example through debriefs). Entrepreneurs tend to need external events for motivation, and as such respond well to requests but not to processes (indeed Attention Deficit Hyperactivity Disorder may be unusually common in entrepreneurs[2]).

It’s also important to look at context. It is typical for progress points to exist at which the NV’s sponsors assess the viability of the company before it can move to the next stage. Sales figures should not be given undue weight here over intangible measures. An important lead indicator is what the Balanced Scorecard calls the ‘learning and growth’ perspective[3], ie, the NVs demonstrated ability to show that it can scale-up its management and practices effectively. The onus on NV sponsors is to make this element clear, and to increase its importance in NV activities. Informally, this is already done when entrepreneurs are asked to produce a business plan that describes not just the idea, but how they will form the company to execute it. Unfortunately, once the company is up and running, these issues tend to get less attention than they deserve.

Managing knowledge flows over the three horizons

The three-horizons model talks not just about one set of horizons, but about embedded horizons too. Thus a business like Unilever sees an NV as a third horizon activity, because it’s a pilot in the context of its portfolio, but to the NV leader he is clearly chiefly involved in horizon-one activities. From a KM perspective, the parent organisation will be keen to learn from the NV as a prototype for what could become a core business (a ‘spin-in’).

Though many NVs are initiated for the learning they will provide, it is less clear whether such learning eventually percolates back into the parent operation. To make it easier to replicate in another area, the organisation needs to know not just the formula for the products, but how the business was set up and grown. NVs cannot simply be shrink-wrapped as best practice, but they can be replicated if you know the basis on which decisions were made, and can therefore adapt them to new circumstances.

The versioning approach outlined above provides much of the necessary knowledge capture, but don’t underestimate the need for tacit knowledge transfer as well. Nothing is more powerful than having the entrepreneur on hand to consult; versioning just helps him remember the nuances of the decision (all of this is similar to franchising – see the separate text box elsewhere on this page). Tacit knowledge transfer can be achieved through apprenticeships and shadowing techniques; think of how people’s jobs might also move through three horizons as they innovate, refine and eventually standardise what they do.

Summary

New ventures are emerging knowledge ecologies that need careful handling if they are to prosper. KM can help a small company evolve through the early growth phase where it is the quality of the management, rather than the business opportunity, that is tested. Some key points to remember:

To help manage the extended team, knowledge mapping techniques can catalyse knowledge sharing and highlight risks;
To ensure that the knowledge base grows in step with the new organisation, establish a ritual of capturing the context of decisions through versioning;
To ensure the option to replicate or spin in the NV at a later date, investors need to create the right environment by emphasising organisational learning.

Acknowledgements

The versioning work described here was designed and implemented with my Unilever colleagues Cees van Elst and Simon Masterton. Thanks to Georg von Krogh for bringing versioning to our attention, Bert van Wegen and John Barber for useful discussions, and to HLL for its co-operation in this research: Balu Iyer, Gaurav Bakshi and K.S. Srinivasamurthy.

References

1. Baghai, M., Coley, S. & White, D., The Alchemy of Growth (Perseus Publishing, 1999)

2. Levander, A. & Raccuia, I., Entrepreneurial Profiling (www.sses.se)

3. Kaplan, S. & Norton, D., The Balanced Scorecard (HBS Press, 1996)

4. Cope, M., Know Your Value (Financial Times/Prentice Hall, 2000)

Sam Marshall is a knowledge management specialist for Unilever R&D. He can be contacted at sam.marshall@unilever.com

Textbox 1 – Why traditional KM won’t work

NVs are clearly knowledge-rich areas, so KM ought to have something to say. However, there are a number of mismatches:

Underlying models – most KM literature takes a diagnostic view, assuming that something has gone wrong that KM tools can assess and fix. With a new company, nothing yet exists to repair, so the challenge is about designing-in good KM practice from the start;
Scale – many KM interventions are about fixing communication barriers that arise with scale. NVs are usually a handful of people that meet regularly, so this isn’t really a major barrier. If anything, scaling-up personal knowledge management seems more appropriate[4];
Resources – even if existing KM were suitable for small operations, the cost of most dedicated solutions would be prohibitive.

Textbox 2 – Sony and Nintendo: knowledge risks in outsourcing

In the early 1990s, Nintendo was videogames market leader. Sega, its main competitor, caused panic by releasing an add-on to its rival console product, using the then fledgling CD-Rom technology. Nintendo approached Sony to develop a rival system. Sony (which had been quietly watching the console market) eagerly accepted, seeing an opportunity to develop a knowledge base at Nintendo’s expense. Nintendo stopped the partnership only when it realised that Sony had market intentions of its own. By then it was too late: Sony had developed the necessary expertise and the Sony PlayStation went on to sell over 100 million units, dominating the market for the next five years.

Textbox 3 – Pierre Victoire: managing knowledge over three horizons

Pierre Levicky was an Edinburgh chef with a winning formula: decent French cooking at ultra-keen prices. He expanded rapidly by franchising. Aspiring restaurateurs received Levicky’s cookery training and he provided them with new recipes each week. Brand awareness was high, and in seven years over 200 branches had sprung up. By 1998, he had gone into receivership.

Rapid expansion meant that new chefs hadn’t really gained the knowledge they needed. Not all of Levicky’s experience could be packaged and scaled so readily. Quality became variable and the brand deteriorated. When some of Levicky’s other investments failed, the whole Pierre Victoire franchise chain came crashing down.