Your service team is full. The teams behind it say they are still interrupted too often. That looks like a staffing problem and it rarely is. It is what happens when the place where the question arrives is allowed to decide nothing. This article sets out where that set-up comes from, how you count in one week how much work the set-up itself causes, and the three questions with which you look again at the design of your first line.
What fits, goes through. The rest waits, or finds a way around.
What you take from this
Centralised or decentralised is the wrong question. What matters is where the room to decide sits in relation to where the question arrives.
How much work your own set-up causes, you can count in one week. Further down is how.
Different forms exist, and the difference sits in who owns a question the answer to which does not exist yet.
What this is about
The set-up is often the same. One service team takes in all incoming questions, and whatever is not solved there goes to the second line. What we call a customer here can be a real customer or a colleague from another department who needs something to get on. The advantage shows straight away. The teams behind it are relieved, get fewer interruptions and keep more time for their own work.
And then something shifts. The first line may only offer what has already been worked out. For everything new or composite, ownership sits at the back. The only room left is the length of the conversation. Help someone on as fast as possible and move to the next.
That split was designed once
It did not come about on its own. It is described in Where Does the Customer Fit in a Service Operation?, an article by Richard Chase in Harvard Business Review from 1978. His proposal was to separate work with customer contact from work without it. That second part could then run like a factory, with batch planning, work measurement and if need be a cheaper location. The contact part went to people who are good with people.
That design is almost fifty years old and it was partly right. For questions that repeat and whose answer is settled, that separation delivers what it promises. It is not a bad design. It is a design with one condition: the questions coming in have to resemble each other enough. That condition is rarely said out loud.
Today it often stands the other way round. The factory has been moved to the front, to exactly the place where the customer contact sits, with call lengths, scripts and targets per call. And the separation no longer runs between contact and no contact, but between simple and complicated. What is simple stays at the front, what is complicated goes to the back. With that, the one condition drifts further out of sight.
What that cut costs
A service team can only answer what comes in if it has enough ways to answer itself.
If a hundred kinds of question come in and twenty answers are ready, eighty are left lying.
Separating taking in from solving is not a problem in itself. A first line can take in questions and solve them while specialists carry out part of the work. It goes wrong when ownership travels along with that handover, because then they see the problem at the front and may do something about it at the back.
As long as the questions resemble each other, you notice nothing, and at the start they do resemble each other. But options are added, exceptions, new systems and new agreements, and each of those drives the questions further apart. The first line does not change with them, so there are more and more questions it cannot answer straight away.
None of this is new. It is the heart of sociotechnical design, the school of thought that grew large in the Netherlands and Flanders with the work of Ulbo de Sitter. You do not have to know it to see what it costs, because you can count that.
What you can count
Two kinds of question come in. One is simply the work. The other is only there because something earlier did not happen, or did not happen well. Someone calling back. A ticket reopening. Someone who after a first attempt tries again through another channel.
That second kind costs you people and hours and gives nobody anything. With a first line that may only give answers already worked out, there are usually many of them. Count them for a week, without the names, and you have a figure instead of a discussion.
At a contact
Did this person already call this week about the same thing, without us having agreed on that?
Is it about something that seemed to be solved already?
Is this person coming to us because there was no answer somewhere else?
One yes is enough. After one week you have a percentage to start from.
When one team becomes the customer of another
There is one more effect, and it is about relationships. As soon as there are two teams with a handover between them, one grows into a supplier and the other starts behaving like a customer. That is a second customer, an internal one, and it should not be there. It is not a matter of character either. It is what a fixed handover does to people.
You hear it in the language. Pass it on. Send it back. Not for us. Every team guards its own agreements, while the only agreement that counts is the one with the person who asked the question.
Forms we often see
Below are three of them, each with advantages and disadvantages, and there are more. What tells them apart is one question. Who owns a question the answer to which does not exist yet?
Owner at the front
Central, with a mandate
One service team takes in all questions and stays the owner until the customer has been helped, even when the answer is written down nowhere. The other departments then deliver to that team, not to the customer. Being the owner does not mean the team solves everything itself, it means it keeps progress and contact with the customer in its own hands. This only works with that mandate. Without a mandate, central is a queue with a friendly voice.
Owner with the team
Decentralised
There is no separate first line. The team doing the work also takes in the question, and knows the customer, the language, the product and the service well enough to answer composite questions straight away. This works when the questions differ a lot and the answer is hard to write down. The price is that every team gets interrupted.
Owner at the back
The in-between form
A team hands out only its settled questions and puts itself in the second line. The difference with central is not in the staffing but in the ownership. Here the team stays the owner of everything for which no answer exists yet, with central that sits at the front instead. This works when the part handed out is genuinely settled and the team really picks up the rest. If that last part does not happen, you have the set-up from the start of this article back again.
There is no good and no bad form here, and in practice it is often a mix. So the design work does not start with picking from this list. It starts with what is not working well in your organisation today. From that follows which form answers it, whether the condition that comes with it holds for you, and whether you can live with its price.
How this runs in practice
Two examples from our work, each with the cause on a different side.
Case 1 · Marketing and communication, retail
Too much on one person
Each brand had one fixed point of contact, and that person had to know everything about that brand and about what marketing and communication can deliver. Sometimes too little work, sometimes far too much, and nobody could step in. We spread the brand knowledge and the specialisations across the team, and split the intake into small questions and larger assignments.
What that delivered. Peaks and troughs are now absorbed by the team instead of landing on one person. New colleagues start inside that team, with coaching and time to grow. Less dropout and more joy in the work.
Case 2 · Internal IT service, public sector
The second line did not get to its own work
One application manager did everything, from first line to project work, and because of the customer contact the second line did not get to its own work. We set up the servicedesk with the second line jointly responsible for the quality of the first. Recurring incidents went from then on to the application managers instead of to the manager.
What that delivered. Clarity, and a better relationship with the business. The servicedesk now saw all the questions from the organisation together and could count how often something was about the same device or the same product. With that it came with facts instead of impressions, and with proposals to improve exactly the things most questions were about.
In both cases the solution did not lie in more people at the front. It lay in who has to know what and who may decide.
Three questions that help set a direction
1
How different are the questions, and do you want to engage with them?
How different the questions are, you can measure. That does not depend on how many departments you serve but on how much you can describe in advance. Whether you engage with the rest is your choice. If your organisation buys Android devices together, may somebody ask for an Apple? No gives focus and lower cost, engaging gives customer experience. Both are possible. It is what you want to be as an organisation, and the whole design hangs on it: which services there are, how they relate, and what each of them is responsible for.
2
How much of the knowledge can be written down?
What you can write down, you can move to the front. For questions for which no answer has been described yet, more than a procedure is needed. Sometimes that is one person who knows the trade and may decide something. For composite questions it is a group that works out an answer and a solution together.
3
What does a handover cost the customer?
A handover that takes a minute with you can mean a week of standstill for the customer. That bill does not stay with them. They solve it themselves outside your service team, or they ask somebody else, or it runs through their manager to yours. What you saved on the handover, you pay back in irritation, in extra work and in conversations between managers that should have been about something else. In the figures of your service team you see none of it. In those of your organisation you do.
When this helps
This helps when your service team is full while the teams behind it say they are interrupted too often. When the first line is friendly and the customer still has to call twice. And above all when you are about to set up or expand a first line. Ask first what that team may decide itself, because that answer decides what the rest of the design is worth.
And it is about more than a servicedesk, and about more than IT. A handover is not a process step. It is a choice about who may decide something, and you make that choice everywhere work goes from one team to another. In HR, where a question about a contract or a hire arrives with someone who may not answer it. With a shared service and the departments it serves. Between business and IT. The same three questions apply there.
And the people in that team
Two things decide how work feels. What is asked of someone, and how much room that person has to arrange it themselves. A lot asked with that room alongside is work people grow in. A lot asked without that room is something else, and nobody keeps that up for long.
A first line that may only give answers already worked out sits, by design, in that second box. With the retail group above it was the other way round, with too much asked of one person, and that costs just as much.
That is not a side effect you make up for later with some attention to wellbeing. It is a consequence of the cut, and it can be solved in the design.
That is why this comes first with us and not last. We design work in which people can contribute to the goals of the team and of the organisation from their own role. What the organisation gets out of it and what the people get out of it stand side by side here.
What the strategy does and does not decide
If customer experience comes first in your strategy, you need experts who sit close to the question, and the first line and the experts work as one whole. If cost control comes first, a separate first line comes into view, but not automatically, because even then the question remains what you do with questions for which no answer exists yet. So that order weighs in. It does not decide. What decides is the relation between where the questions arrive and where the room sits to answer them.
How we go about it
Workable work is a design requirement for us and not a wish afterwards. From there we start with your strategic plan and with the question as it arrives today, not with a model for service organisations. From that follows how different the questions arriving at your service team are, and how much room it needs to answer them.
We make the design with a group from your own people, because whoever has to carry a change helps design it. We do not fill roles and we do not take work over. We stop when the new way of working is the ordinary way of working. Read more about how we work →
What may your first line decide itself when the answer is written down nowhere?
In a conversation of 30 minutes we walk through the three questions for your own service team. Afterwards you have a picture of where the room is missing, and a way to count that the week after. Whether there is a follow-up, you decide later.