Wednesday, 18 February 2009

Service Innovation Part 4: Safe-fail experiments

Although not exclusively applicable to services, experimentation plays a vital role in the development of new services. 

Traditionally however, experiments were associated with Labs, where people - often nerdy white-coated chaps (and few girls) - built heaps of new knowledge doing interesting things with stuff. These days are not completely gone, but for the sake of the argument, lets says that the golden days of this approach are behind us now.

Contemporary services literature in general and the services marketing and services design literature in particular approach services either as an extention of product (and can thus be designed in advance) or as innovation that somewhat magically emerges from customer interaction. I believe that both perspectives are fundamentally flawed for two reasons:
  1. Although a lot of effort is put into "user-interaction design" when developing a product, still most users must spend considerable effort before they feel "at home" when using the product. Towards the services-end of the spectrum it becomes less and less possible to design the user interaction as the output of the service becomes more and more determined by the customer/client/user. Any up-front design activities will limit both the services-recipient and services-provider in their flexibility to co-create value.
  2. In daily services operations, unexpected customer requests, or other non-plannable circumstances are a fertile source of ad-hoc experimentation. Quite often best efforts of all involved lead to good (and regularly co-created) solutions that satisfy the urgent needs. This way real-time, ad-hoc and to a lesser extend on-demand novel practices emerge on a daily basis in the front-stage. But also on a less heroic level projects that deliver customer-specific systems (products, services, hardware, software, etc.) are in fact often producing novelty on a regular basis by going from one customer to the other and fror one context to the other.
The lattter form of "creation" in my view is one of the most fertile source of innovations. Things that work for customer #1 are varied upon for customer #2, #3, #4 and maybe than #1 again, thereby creating a group of costomers that each have a somewhat similar/somewhat unique incarnation of a system for which no marketing material exists.  This way of creating an installed base in fact "innovates" a group of customers one by one, learning-better while doing. 

From a developers perspective each project in the series can also be seen as an experiment to develop a better version of "the solution". But since the customer pays, it is also quite safe to do, because one failed project only harms one customer, not all. This is one of the principles behind the Safe-Fail approach: multiple tries where one failure doesn't disrupt the whole while variation between the experiments increases the chances to achieve/find/stumble upon real breakthroughs. This is how a large part of the creative/implementative part of services industries works.

(to be continued)

Monday, 16 February 2009

Service Innovation Part 4: KEBS

After "clusters" and the "frontstage" today I want to introduce you to one my favourite new terms: KEBS. Those familiar with the Services Innovation literature might suspect a typo since KIBS is a well-known acronym in those circles.

KIBS stands for Knowledge-Intensive Business Services, i.e., Services that require highly-educated people: architects, consultants, laywers, medical doctors, etc. Some of these expertises work in solo groups. Architects for example can do their "thing" from relative isolation, but the medics are surrounded daily by scores of people in more or less supportive roles: cleaning, nursing, labo-testing, catering, etc.

That kind of work is what I call KEBS: Knowledge Extensive Business Services. And KEBS are LARGE - no even HUGE - in economic terms. Back-of-the-envelope calculations easily show that besides the 20% in Manufacturing and Agriculture, less than 10% of the economy is KIBS. For example, have a look at these numbers from the Dutch Bureau of statistics:

Probably illegible (and in Dutch), but on the left are the largest services sectors of the Dutch economy in 2000: wholesale, healthcare, construction, trade, banking, legal services, employment agencies and post/telecommunication are the only sectors that score more than 2% of the BBP (Bruto Internal Product).

It might be a somewhat naïve view, but of these 8 sectors only healthcare, banking and legal have substantial KIBS activities, but even there large chunks can easily be described as KEBS: nursing in healthcare, administrative processes in banks and dossier management in Legal are clear KEBS examples, just like cleaning, facility management and mail processing in all three. Overal my estimate is that KEBS represents over 90% of the 80% that we call the services economy, that is >70% of our economy!

This silent majority is the main target for service innovation in my view. How to facilitate nurses to a an even better job? How to bring about more cooperations between healthcare and wellfare silo's to improve the service to citizens? And lets not forget the governments too! How to improve the functioning of all those public bodies? These are the real Services Innovation/SSME topics that need to be adressed in the forthcoming years. And solutions need to be implemented in those overwhelmingly large KEBS based activity systems.
So that is my next favourite acronym, KIBS-SI-KEBS-S:
Knowledge Intensive Business Services for Services Innovation in Knowledge Extensive Business Services Systems
Wow, what a bullshit-bingo term! But for the moment I don't have better one. Suggestions?

Saturday, 14 February 2009

Service Innovation Part 3a: Frontstage Tools

At times Lady Fortune seems to exist. Just after finishing my previous blogpost I learned that the Civil Services in Singapore has started a narrative sensemaking project to better understand their employees' perception of working in the Civil Services

So what is going on there? Well, we are all used to surveys to dig deeping into problems. Questions like "how would you rate our service" or "please indicate which problems/items belong to you as a Civil Servant" are routinely asked. Such an approach assumes the researcher is supposed to know what problems to look for. He/she has probably formulated an hypothesis or a set of hypotheses of possible problems in the target group.

And no wonder - lo and behold - this is what the result will look like too. Some hypothesis are proven, others are discarded, thereby  supposedly confirming or discarding the hypothesis, often accompanied by some statistical proof about the validity of the results. But what is the hypotheses are wrong upfront. What if there are other issues in the target group. What about weak signals (that only a few respondends fill in under the "what else do you want to share with us box") that are signored because that box is hard to interpret or suppressed by the statistical methods used to analyse the results (that in general look of majorities, not minorities)?

In such cases (and in my view this is always the case), one should use research methods that are in the pre-hypothesis class: no up-front problem list, no upfront filtering.

This is the approach taken by the Singapore Civil Service. They state:

This project will collect stories from a wide variety of College and other civil service staff, capturing their perspectives on work life within their organisations. The project is expected to reveal insights for new initiatives that influence staff engagement and will provide deeper meaning to the existing climate surveys already in place.

And that is how it works. These methods are attached to an existing practice of surveying to "provide deeper meaning" and "reveal insights for new initiatives". Both are achieved because pre-hypethesis research methods always easily surface new issues that surveys cannot pickup by design and they also provide real-time insight into how people as a collective respond to developments, even if these are from outside the organisation. 

Next, trough a process of emergence, patterns are formed from the narrative input, thereby producting of meaningful results in which the weak signals are showing up too.

If you want to know more about how such methods can be applied in the front-stage of your organization, please have a look at the Cognitive Edge website or contact me for more information.

Friday, 13 February 2009

Service Innovation Part 3: Frontstage

In my previous post on this subject I have first given definition for both service and innovation in Part 1 and gave my take on the breakdown of the services economy in Part 2. I argued that we should forego the idea of economic sectors and instead focus on clusters of similar activities and we should also give up the product/service devide and use a transaction/transfer perspective instead.

Besides Araujo and Spring another source for the same idea is the book Services is Front Stage - We are all in services ... more or less by James Teboul. There is a lot of interesting stuff hiden in that book, but one of the most clarifying is the picture on page 8:

In it Teboul compares the customer-facing activities (front-stage) and production-like activities (back-stage) of three types of restaurants. First McDonalds on the left. Their backstage is enormous. The raise pigs, convert them into hamburgers in their own factories, run their own distribution fleet, are famous for their minite kitchen operations and run their business is very standardized outlets. That is the backstage. The customer facing part of McDonalds is quite meagre. A small front desk with trained employees that are all equally friendly, serving standard meals on trays, with self-seating in the "restaurant" where also the famous "ball and slide area" is situated and that must be looked over by the parents.

In  the middle we find a gourmet restaurant that has a kitchen that buys its own stuff from local sources, that prepares meals on order and that can handle special wishes like "done" or "half-done" or "bacon with egg, but please replace bacon with ham". That is the backstage, a lot smaller than in the McDonalds case. The frontoffice is quite big. Often it is a nicely fitted area with waiters that know their job, can advice customers on dish/wine combinations.

On the right we find the Benihana type restaurant. To my knowledge an unkown type in Europe, but maybe it best compares with the Japanese style restaurant. The backstage is very small in that type of restaurants. Food is sourced from "the best" sources and preparation is little as most of the work is done by chefs that prepare the food on the table, in full sight of the customers. They do their work in a way that we can truly speak of a dining experience.

Overlooking the three types of restaurant we see that the backstage is getting smaller and less organized, and the front-stage is getting bigger, but also better equipped with staff, luxury and increased dining experience. It will come as no surprise that Teboul argues that all three types of restaurants are in services, but for the Japanese/Benihana type the services part is much much bigger than McDonalds'

To me the front-stage/back-stage example nicely demonstrates the economic clustering idea.

Tuesday, 10 February 2009

Service Innovation Part 2b: Clusters revisited

Oops, I didn't write a word on this blog for almost two weeks now. This way this is more a column than a blog.

And it seems I'm not the only one. Colleague blogger Samuel Driessen seems to be quilty of the same sin. Is it our location (Netherlands), it is the economy, Obama politics, shared employer, the weather? No clue, but in my case my unhappyness with my previous blog clearly added to the delay. For a while however I couldn't figure out what was wrong with it, but last weekend I found out where the bug was.

The title was OK, economic clusters. But somewere halfway I changed the subject to transactions and transfers as economic entities. So where did the cluster part go? Well, nowhere, so here it is. Finally.

The point I was trying to make by citing the Araujo/Spring article was basically that products are not transactional in economics terms when they are not accompanied with or even covered in some form or services such as warrenties, retailing, installation, insurance, after-sales services, etc. And when combined both physical and non-physical products can be well described as transactionable goods. So far, so good.

But, economists clearly classify the insurance, retail, installation and construction industries as services industries leaving only manufacturing and agriculture as product industries. Hence the rapid growth of the services percentage in the GNP figures. Araujo and Spring offer another alternative by focussing on the transactionality of the economic entity thereby effectively realising that a car with warrenties,and a dealer network becomes a transactionalable entity just like 8Mbit ADSL for one month from Vodafone. So Goods Rule!

But what to make of this example - a little story ....

I already told about my visit to the Frontiers of Services conference in San Fransisco last year. One day I was listening to a lecture by Walter Ganz from Fraunhofer Institute in Germany and he asked the following question: suppose BMW decides to outsource its canteen to Sodexho, what will happen to the services share of the economy? Before the outsourcing the canteen is figured into the numbers of the manufacturing industry, and afterwards it is in the catering industry which is accounted as services! Meanwhile, Walter stated rightly, nothing changed!

These two points (goods rule! and nothing changes through outsourcing) are the key reasons I believe we should stop accounting our economy by sectors:
  1. By looking at the transactionabililty of an economic activity we can overcome the difficulties associated with the product/services dichotomy. Instead we can focus on transactionable goods and transferable services. This will largely blur away the strict sectoral boundaries between OEM manufacturers and retailers, between insurance/banks and ICT companies and between architects and the building industry. It think it would be fairly easy (but still a formidable task) to break our economical statistics into transactionable goods and transferable services. If we could break down the revenues and profits of the current sectors into these two numbers a lot of the confusion raised by the product/service classification will be cleared.
  2. If we would to the same thing Dave Snowden suggests to do for narrative and clusters the activities of organisational entities on a wide range of opposing scales (for example efficiency focussed operational versus freaky controlled managerial) we would obtain a multidimensional dataset in which clusters of similar economic patterns (for example maintenance, selling, displaying, giving advice, helping achieve, counselling) will appear that were previously spread over multiple industries. It will also turn out that the same activity will be part of multiple clusters when displayed using different axes. And indeed, work done on a factoryline is both a production activity and a way to have social interaction with colleagues. This way we will enable ourselves to have multiple perspectives of the same activity which will also greatly enhance our understanding of what is really going on in our economy.
Well, I guess this one is covered, finally.