- Creating a Process-Centric Organization
Creating a Process-Centric Organization
Operations Director Graham Underwood set up GFT in the UK, growing the turnover from 0 to £50 million in ten years. In the three years to 2015, he took the firm from 1,300 to 4,000 staff. He talks to Steve Brooks about the importance of creating “a process-centric organization”. For Underwood, making sure the processes are right is a first step.
He pays particular attention to those that run across the business, the lead-to-order process, the order-to-delivery process and the month-end close. The key to successful change is senior champions, Underwood explains.
Doug D’Argenio: Welcome to the PS Insights podcast series, sponsored by Kimble Applications, professional services organization, strive for efficiency, success, and growth. This series is intended to provide key insights on how to achieve this from industry leaders.
Steve Brooks: Hello, my name is Steve Brooks, and today I’m talking to Graham Underwood about creating a process-centric organization. Graham has worked in professional services for more than 30 years. He set up and was Managing Director of GFT UK, a professional services firm operating in the finance sector. As Group Operations Director at GFT he helped grow the company, increasing revenue over three years from 121 million Euros to 374 million Euros, and from 1,300 to 4,000 employees. He now works as an interim Operations Director or Programme Manager. Graham hello.
Graham Underwood: Good morning.
Steve Brooks: When you’re talking about setting up a process-centric organization, when should you first think about defining those processes?
Graham Underwood: Well I think the simple answer is early. I think you need to, first of all, sort out your processes before you build software. The worst possible thing you can do is start automation of bad process or with no clear opinion of what the processes are.
Steve Brooks: But some software will say that they have all the best processes already for your organization, so should you not look to bring that software and just run your business around that software?
Graham Underwood: I think it’s a balance. I think you’ve got to understand what your processes are to start with, but yes, if you’re implementing particularly some very standard processes, maybe some back office processes, you should be prepared to adapt your processes to a product because it’s much more efficient to tune the processes and use the product as it was designed to be built, than it is to try and change a product. That leads to ongoing maintenance problems, ongoing cost in the long-term, so you have to be flexible.
Steve Brooks: What are the important areas where professional services firms need to think carefully about its processes? And also, who should actually start defining those processes?
Graham Underwood: It depends, you can either start looking at your core processes or your support processes. It’s very easy to look at support processes, let’s for example talk about recruitment, or talk about invoicing, but there the processes which are very important to a professional services organization, but I wouldn’t say they’re core to the organization. The core processes would be, for example, lead to order process. In other words, what’s the process for, finding opportunity, and converting that into a piece of business. Or look into a process of order to delivery, what is the steps that you need to take once you want a piece of work to actually deliver it to get client satisfaction, to make sure you invoice the client correctly, and do that on time and efficiently.
Steve Brooks: You grew GFT enormously during your tenure there, what were the processes that really made a difference during that period as you scaled the company up?
Graham Underwood: Well GFT grew a lot through M&A and when you do that you bring in companies with very different ways of doing things. One of the core key processes that we discovered that was slowing us down was the month end process, which is the process of collecting all the time that people have put onto projects, getting all the milestones that need to be invoiced, doing the invoicing and recognizing the revenue and reporting back to the markets, which for a public equity company has to be done in a timely fashion.
Steve Brooks: When you’re doing M&A work, you obviously bring in companies from elsewhere which have existing processes. Did you ever look at their processes and steal their processes to come into GFT?
Graham Underwood: Yes, we did. It’s important when you’re doing M&A, it’s a joining of two companies and best practice might exist in the company that’s being acquired. Yes, we definitely looked at the processes in the other organizations where they were better than the ones that we already had we would adapt. Having said that, of course GFT was a publicly quoted company and certain processes like the finance processes were very robust and they’re not ones that you would change.
Steve Brooks: Having seen lots of processes during your time there, what’s a good process look like?
Graham Underwood: Firstly, a good process is homogeneous, and homogeneous is the very important aspect when you’re growing a company certainly at the speed that we were doing at GFT. By homogeneous it means everyone does the same thing the same way. That’s very difficult because there will always be reasons why, for example, in certain countries, in certain geographies, things have to be done different. There might be tax reasons or there might be legal reasons in a particular country. The important thing is that you look to standardize the process and treat the specials like a special tax treatment as a one off in a particular country, you don’t develop a whole new process for something just because one part of the process is different in one of the countries. The phrase I like to use was lowest common denominator rather than highest common multiple in effect, always look for the common pieces of process and make those robust, make those standards across the organization. Don’t try and build a process with every variable for every different country.
Steve Brooks: You’re touching on change management there. Once you’ve established what your best practice process is how do you actually roll that out? You got any tips for rolling it out at all?
Graham Underwood: Yes absolutely, the most important thing is to have process champions. Without an owner of the process, someone who’s going to take responsibility for it, someone who’s going to develop KPIs and measure the process, somebody who’s going to ensure people are trained to run the process, someone who ensures that there’s the right capacity to execute the process, you’re duped. Without a champion, and this champion needs to be relatively senior in the organization because they need the gravitas to make it happen. Without the senior champion you’re lost. Now the challenge that I’ve observed is that the senior people are also awesomely busy and therefore it’s very difficult to get the sufficient bandwidth in the process organization to successfully deliver it. That comes down to leadership. The process champions need to be C level members of the organization, maybe the top-level person responsible for the processes is a board member or certainly reports into the board.
Steve Brooks: Is there risk with those champions becoming bottlenecks?
Graham Underwood: Absolutely, yes. As I said, the champions tend to also have other high-profile jobs in the organization. For example, let’s take sales, a senior person in sales his primary focus and his primary value to the company is to generate revenue. And therefore, to have to do it or responsible for processes as well is a conflict.
Steve Brooks: You’ve got these champions, but surely if you’ve got a champion owning a process, doesn’t that sometimes mean they actually become a bottleneck for the process? How do you define the process so they’re not a bottleneck?
Graham Underwood: You bring up an important point here. If I could bring up an analogy with software development. I think it’s well understood, an industry norm that project management you cannot rely on heroes. There are many standardized approaches to delivering projects, for example, Prince where you follow a standard way of delivering a project and it means you don’t have a single point of failure. You haven’t got a hero working 80 hours a week trying to deliver a project. I’ve seen it happen so many times, particularly in internal projects at companies where people work crazy hours and think they’re doing a great job for the company, but then if that person goes sick, which can happen if people are being stretched that much, the whole project delivery can break. And you get the same in a process organization. If a key hero running a process, one person, a single point of failure is a disaster. If that person leaves the company or goes sick the process can stop, and if that’s a mission critical process you’ve got a problem.
Steve Brooks: In things like lead toward a process you talked about earlier, that’s not just impacting sales department, it also impacts delivery and operations departments as well. So where does the champion fit within that situation? Are they head of sales? Are they head of operations?
Graham Underwood: This is the challenge. Most organizations that I’m aware of, as they grow, they tend to grow in a siloed way. So, particularly in the support functions, the support processes. So, there will be a finance department run by a CFO and he can, within his organization, can create very efficient, very structured finance processes. Similarly, in HR, very structured efficient, recruitment process or appraisal processes in a professional services organization. The challenge is that the end to end business processes cross departments, so the process you talked about lead to order, it would cross sales, it would cross delivery, it would cross finance, it would certainly cross HR, you’ve got a staffing implication. And finding the champion that can own that end to end process, where it potentially comes into conflict with the nice efficient processes that the CFO has designed to run his finance department efficiently creates a conflict. That’s why I’m saying that the champions of these processes need to be senior level people, they have to be able to operate at board level, and they have to be able to negotiate and work within a matrix organization in order to achieve efficient processes.
Steve Brooks: Effectively what you’re saying is, taking the lead to order again is the finance team should not own that operation because then it’s finance centric, not necessarily business centric.
Graham Underwood: Yes and no, it depends on the maturity of the organization. A more mature process organization will be focused on the end to end processes. A less mature organization will be focused much more on departmental process, then the end to end processes suffer. This comes down to governance. We haven’t touched on governance yet. But changing a process requires the cooperation of process owners who are responsible for an end to end process, as well as a process owner for a department. For example, if the finance department decides to changing their invoicing process without consulting end to end process owners, the end to end process might break and I’ve seen that in organizations.
Steve Brooks: Can you give an example of where it’s gone disastrously wrong?
Graham Underwood: I was running an integration programme, joining two companies together, and I had a global team spread across different countries, and different parts of the team were responsible for different parts of the process. What I didn’t realize that in our organization that the way the finance processes were run in one country were completely different to the way the finance processes were run in the other country. So, both parts of the team worked on bits of this integration programme not knowing that they were doing different things. We’re only a few weeks away from go live point where we had to launch, the end of the financial year, we had to finish the integration. We had this oh my goodness it’s not going to work moment, when we realized that the invoicing process was broken because half the team had designed it one way, half the team had designed it the other way, it was a real crisis. And that was the trigger point in that organization to say we really now need to sort out our end to end process, look for homogeneous processes, we literally cannot grow the company anymore until those problems are solved.
Steve Brooks: As companies grow and especially internationalize, that’s almost a key point where organizations or leadership teams need to be aware of their homogeneous process across the whole organization, is that fair to say?
Graham Underwood: Yes, I think companies reach a tipping point. They hit the wall where reaching a certain size calls for having structured processes across the company, but also having structured processes across the company becomes a prerequisite for further growth. In a small company the management are able to oversee everything as a company’s growing. The founders, they’ve built a company from scratch perhaps, they know how it runs, they know how everything works, they can oversee everything, but the tipping point is when the company gets to the organization that the small number of founders, they reach a point where they cannot manage everything to sufficient detail.
Steve Brooks: Is it fair to say that those founders should be considering the processes before they reach the tipping point, and actually if they’re strategic intention is to grow the company rapidly, they need to make sure their organization is structured and has the processes ready for that growth ahead of the point where they reach that growth?
Graham Underwood: Yes absolutely, if you look at the statistics, 90% of starter companies are unable to scale and ultimately fail.
Steve Brooks: When you’re looking to internationalize is it to build multi-national processes, and are there any specific process areas that can cause real difficulties?
Graham Underwood: That’s an interesting question. I think there are two types of global company. There are international companies and there are multi-national companies, and it’s a subtle difference. A multi-national company is one where you’ve got operations in all over the globe, which may have been organically or through M&A, but you accept that those companies run themselves and you’d rather consolidate the numbers at a group level. An international company is one where you do strive to integrate processes globally, so everybody does things the same way and you have one set of processes governing the organization. And you can run an organization both ways.
In an international professional services organisation, where perhaps people in one country are doing delivery work for people in another country, as typically happens in the financial services industry for cost arbitrage reasons, where you’ve got people in low-cost locations like Asia or South America, delivering work into the banks in Europe or North America. You’re managing a pool of people globally, and therefore you have to have processes around the core processes that run the organization, which are harmonized. For example, you’ve got to close all the time capturing for the month in order to understand the revenue that you’ve generated for that month. You’ve got to close that globally, and that’s where tools can come in because having a platform in which you’re running your professional services organization can put in place the discipline and structure to force that.
Steve Brooks: We’ve spoken about bad processes, are there any warning signs that you’ve got a poor process? You talked briefly about the heroes and the bottlenecks they can create, are there any other bottlenecks that you need to be aware of?
Graham Underwood: Well firstly, you need to be able to measure your processes, which comes back to the process champions. You need to have KPIs, you need to be able to see how your processes are performing, and be prepared to make change. I think there’s two sorts of bottlenecks. I would call them short-term and long-term bottlenecks. A short-term bottleneck would be a temporary problem, and a good example from that would be, again coming back to the hero, you got one person doing a particular task in a process and that person goes sick and suddenly the process stops because you haven’t trained backups, you haven’t scaled the organization properly.
Conversely on a long-term process typically you would have many steps, maybe executed serially, and you’d have to look at, for example, was work backing up in certain places in the process flow, in a long running process? Could it be that people are waiting, they’re not able to do their work before someone else has completed their work, and therefore you could start to look at how can things be paralyzed to avoid these bottlenecks? You might observe consequential factors, for example, people going sick with stress, or people working a large amount of overtime at particular parts of the month, which would indicate you’ve got a stressed process.
Steve Brooks: If you’ve been given the responsibility of setting up the processes, say lead to order or order to cash, are there any approaches that you would recommend people to take in creating those processes?
Graham Underwood: Yes, firstly I would say make sure you document the processes properly, which is a big topic. It’s very easy. Using very simple tools like PowerPoint or Visio, to draw some pretty pictures of flow diagrams and define that, say that is my process. If you take that approach, that doesn’t make it easy to identify problems and make changes. So, in my experience the right way to do this is to invest in some sort of tool to build a process model, which means that once you’ve got a process model with all the interconnecting processes, it means you can understand the change that you make and understand the impact on other processes taking the example we referred to earlier, where you’ve got an end to end process which crosses many departmental boundaries, changing a process inside finance, for example.
If the processes are all captured in the process model you can immediately see and understand what other processes might be impacted by that change. Not only that, but you would link the processes also to IT systems, so when you’re making a change in an IT system you can immediately see from the process model which process areas are going to be affected by a potential change in the IT system.
Steve Brooks: If your company’s expertise isn’t in business process modeling, which may be the case, where should they turn to get help and which software? … Are you talking about software or you’re talking about consultancy?
Graham Underwood: There are plenty of companies around that’ll sell you consultancy to do process modeling, depends on the organization size. At a certain size, and we’re talking organizations of the size I’ve worked in. We built an internal business process management team, which was the owner of the process model. They supported the process owners by maintaining the model and governing change to the model. In terms of tools you can spend $1,000 or $50,000 on a tool depending on what you want. In my experience there are very good tools which don’t cost very much money, which allow you to create a process model, and then publish the model in various forms, in web pages, and documents, which can be used to train the organization. But the important point is the model is all in one place, so every time you make a change to a process all the documentation comes for free.
Steve Brooks: In some organisations, I’ve seen often these kinds of horizontal things get passed across the IT department to run. Your CIO ends up being in charge of processes because that’s what the system does. Is that the right approach?
Graham Underwood: It depends. I think the CIO could also be the Chief Process Officer, the CPIO. In a small organization that would be the natural first step to make. However, I don’t see it that way. I think the CPO, the Chief Process Officer, and the CIO are different, they’re peers. The Chief Process Officer is someone who owns all the business processes, who defines how processes will work. The CIO is the person that runs the IT services, keeps the lights on, makes change to IT systems to support processes, and those two people can be in conflict. CIOs job is to run his IT systems efficiently, make sure they don’t fail, whereas the Chief Process Officer might be looking for a competitive advantage by changing a sales process, for example. Those two requirements come into conflict and therefore you need senior people in those roles who can work together to produce the best outcome for the company.
Steve Brooks: Some people will say, leading on from that, that the processes can kill organizations. They become too process centric and actually sometimes you’ll get people saying actually no, we need to be more customer centric, not process centric. Is the right thing a balance between the two or can companies survive on being purely customer centric or purely process centric?
Graham Underwood: Think it depends on the organizations. Most companies which are delivering a service to clients would say they were customer centric I’m sure, but I think there has to be a balance, there has to be a limit. You can, for example, create a very efficient invoicing process which ran your invoicing extremely efficiently. You only sent out your invoices on the first day of the month and they were always in the same format and you had a very lean efficient finance department. That might be fine if you’ve got many clients who will accept that constraint in how they’re going to be invoiced, but let’s imagine you have a business, particularly in professional services, where a significant part of your revenue comes from a single client.
Let’s say you’ve got 30% of your income coming from one customer, and that customer says, “Well I want my invoicing done in a particular way.” Then you’ve got that challenge. Am I going to be completely customer centric and do whatever the client wants, which makes my invoicing process less efficient, or vice versa and I think most companies with significant client would change the processes to suit the client. Would most companies make that change if it was a one-off piece of work for a very small client? Maybe not, so there is a balance there, I think.
Steve Brooks: When you bring exceptions into a robust process the risk surely over time is that you bring in too many exceptions to that robust process, such that a robust process is no longer robust. How do you identify those and who’s responsible for that? Is That your champion? and how does he go about measuring that or identifying whether the process is now being bypassed?
Graham Underwood: Yes, it comes down to having process champions. For every process there has to be a process owner, the person who’s responsible for the governance, and measurement, and improvement of that process. He or she should have KPIs in place which measure a process, so using that example you’ve just given of creep, where processes have been evolved over time to suit particular client needs for example. The picture I have is adding lots more layers on an onion. So, you keep adding another laying on the onion to solve one problem, and then you add another layer on an onion to solve another problem. And it’s analogous to what happens if you do that with software. If you keep patching bits of software to fix a particular problem, eventually your entire platform becomes unusable and unmaintainable, and the same happens with processes.
So, the process owners need to have developed KPIs, key measures that for their processes. Let’s take a very simple process, a month end process in a professional services organization. What are the KPIs? It might be time if it’s a publicly quoted company you have to report to the markets at a certain time after month end and therefore time is one of the key KPIs. You have to complete that process in X number of days. The process owner must be monitoring that. He must be looking for where the inefficiencies are, he should be continually trying to make changes to improve the process, to improve the efficiency.
Steve Brooks: In the current world and I hesitate to use the word transformation, but we’re seeing a lot of transformation across different industries. Is there a risk if you become process centric you can’t actually innovate or it hinders your innovation?
Graham Underwood: No, I don’t agree with that. I think if you look at what’s happened in the last 10 years whole new companies have sprung up, which have been driven by redefining processes for competitive advantage. For example, take Uber, what Uber have done is completely redefined the process for how you book and take a taxi. They invented a new business model, they invented a new process. Or going back a bit further Amazon completely changed the processes we use for buying or renting books.
Steve Brooks: Can companies make sure their existing processes aren’t being disrupted by competitors? Is it the process owner’s responsibility to review and to look at their processes every so often?
Graham Underwood: I wouldn’t say every so often, I would say continually. In a mature process centric organization, the process organization is measuring the efficiency of processes using KPIs and continually looking for ways to improve.
Steve Brooks: How much difference can robust processes make to an organization?
Graham Underwood: Let me answer that the other way around. By not having robust processes could be mission critical, could be company threatening. If you’re not able to get your invoices out at the end of the month because your processes are broken, most professional services firms that I know would go out of business quite quickly.
Steve Brooks: How do you identify where you need processes?
Graham Underwood: If we focus on end to end processes, as opposed to departmental processes, I think a typical professional services firm might have five, maximum 10 different processes. You’ve got some processes around generating propositions, so an idea to the proposition or something. You’ve got the processes around sales and marketing which is capturing leads and turning them into orders. You’ve got the processes around taking an order and delivering it, billing and the collecting of money from your clients, so an order to cash process. You’ve got a lot of processes around people, because in professional services organization people is your primary asset, so processes around hiring people, and training people, and appraising people to grow them within your organization. You’ve got processes around reporting, so the record to report process, taking transactional data and producing output. And that was I think I’ve said five or six different process areas, that’s it.
Steve Brooks: If you’re a small start-up professional services firm, what kind of approach do you think they should take?
Graham Underwood: In my opinion it’d be a mistake to think you can just go to an external company who would build all your processes as a consulting project and then disappear. You have to have someone at least transfer into the organization because you have to maintain the process model going forward. Maybe that’s with a permanent ongoing consultative support, but there needs to be some knowledge in-house.
Steve Brooks: Are you saying people just evolve that process or if I’m a start-up firm, what should I do about these core processes? Shall I set them up from scratch? Should I just let them evolve?
Graham Underwood: Yeh, I think if you’re a start-up company taking in the process lead to order. If you’re not using that process for a competitive advantage then you can probably take a fairly standard process, which will be good enough to run your organization. If your organization was, for example in fast moving consumer goods, where your competitive advantage was speed, for example, then you might want to tailor that process because it is the key to your business being successful.
Steve Brooks: Thank you Graham.
You have been listening to one of a series of podcasts to dedicated to sharing best practices for professional services organizations. These can be found on www.kimbleapps.com.