To all those anxious to go digital, to adopt the latest AI, to automate, to find the edgiest algorithm machine learning, this is where to start.
So you want to go digital deploying AI etc. but don’t know where to start? Well, first things first: don’t start by thinking about technology! Think about why you need the technology (other than having an answer to your boss when asked).
You are not just trying to go digital; you are also trying to change the value proposition of a future-ready Procurement function. The function of the future is a tech-enabled, high capability business partner, capable of bringing real market insight to the stakeholders whenever they need it (insight partners), and capable of building networks of suppliers and stakeholders to launch the latest innovative products and services (network partners).
This is really important – technology is going to help enhance and free up from the transactional, through the analytical and eventually to the spot buy and much of sourcing itself, as well as providing insights further down the supply chain to manage risk. In other words digital procurement technology is going to shrink the base of your organization pyramid, but…
There’s a starker reality at work. Do you know any CPO actually doing anything in the area of AI, automation, etc? I know a few US CPOs who have really begun to experiment and have started to automate up to 20% of their sourcing activity, but they are outliers who are 2 years in. IBM’s Watson tool is starting to produce some interesting insights into vendor management and contract assessment. On the whole however, most CPOs I know have this feeling they ought to be doing something, but they are not doing it now.
The complication is the structure of the function, linked to the diversity of procurement activities. The classic category-based function has many category specialists doing all of the procurement cycle. This means that investing in automation of one element (eg analysis) is, incrementally, going to improve the job of everyone in a similar role. That is hard to crystallize as an ROI, but remember: so is slowing down automation!
So, what do you do first?
Get your senior team to recognize that this is a journey, and all of you are going to have to change.
Second, you restructure the function to be more like a cake (with horizontal layers performing different roles) and you narrow the range of activities each individual does (ie you have business partners, sourcing professionals, procurement operations doing analysis, RFX construction, market research, etc). When you have 10 people doing the same, much narrower thing, automating a third of their activities starts to give you the chance to actually free up and strategically leverage three heads, and hence build a business case. There are a range of complications about getting knowledge to flow and where to hand over activities, but on the whole, it works.
In other words, to get to a digitally enabled kite function, you need to build a ‘cake’ first.
Third, there is no doubt in my mind that you also need to experiment. Get your head around the ‘art of the possible’ in digital procurement. If you get the chance to see how to program robotic software apps, do it, even if it relates to an area other than procurement (finance – which has many, many more predictable activities on a cycle – and holds the purse strings – is going to automate before procurement). There is a reasonable chance that the first areas that will be useful may be robot software/robotic process automation, and so you may want to experiment here first.
These three things are where you need to start today. Go forth!!