Pricing Automation: Questions to Ask
“The first rule of any technology used in a business is that automation applied to an efficient operation will magnify the efficiency. The second is that automation applied to an inefficient operation will magnify the inefficiency.” – Bill Gates
Pricing as a function has evolved over the ages, and like any other evolving entity it has embraced the need to complicate. Once the complication has reached a point of no return the obvious question that is born is “how do I make the complication efficient?”
Most companies and pricing experts today believe that the answer to the above question is “automation.” Well we are the ones who could replace thousands of humans from the shop floor with robots and computers—we could very well make use of our competency to replace a bunch of pricing managers! Can we not? Being a pricing manager myself I think I could try to answer. Well of course, yes you can.
However there is a counter question from my end—back in the days when you made the shop-floor ‘people lite’ you had a purpose that was well defined: these folks whom you replaced were performing functions that were easily replicable by machines, and machines are far better than humans since they don’t take sick leaves or unionize. Also depreciation and maintenance expense is lesser than wages. What is the ‘purpose’ that is driving you this time round?
Easy one. Efficiency.
Define it please?
Well, from a pricing perspective efficiency is two dimensional—optimal prices (that would make the P&L statement look good and customers would love) and quick prices. We want to be the first to price and the first to win!
That sounds pretty convincing. You will be the first to price no doubt, but I am a bit skeptical about the winning bit. How much have you been winning so far?
Well quite a bit. But at end of the day you win some and you lose some—isn’t it so?
True. The pricing managers have priced to win sometimes, and sometimes they’ve failed. What do you think led to the failures?
They priced higher than they should have, meaning it was not optimal.
So how do you ensure that the machine takes an optimal decision 100% of times?
I don’t know. That’s what machines are supposed to do. They just need the right inputs and the right price pops out based on the right algorithm.
Unfortunately the right price popping out depends on human factors—right input and right algorithm. These are beyond the scope of the machine. Which humans will you hire for the same?
Specialists. Input specialists and algorithm specialists!
What if these specialists are nothing but your old bunch of pricing managers whom you wanted to replace in the first place?
Getting back to the shop-floor example: what if you realized that you ended up buying a few million dollar worth machinery but had to retain the workforce to run the machines as per requirements? In fact you ended up renting a bigger floor to accommodate the humans and machines?
In your pricing automation journey you may end up in a similar ‘investment trap’ unless you ask the right questions at the very outset:
Question 1 – How do we price today?
Simple question. You need to be sure about the starting point at the very outset. Not just sure, yet you need to be in love with the process. The reason is simple—once you automate, this is the process that is going to magnify.
Question 2 – Am I not happy with today’s pricing process?
You should not mix up terminology. If you are dissatisfied with the existing pricing process and want to change, you need to change not automate.
For example your company has been using “cost-plus” technique for ages, which has led to millions of dollars left at the table. As a measure to seal your leaking wallet you want to adopt “value-based pricing,” or something else that’s more profitable. You need to transform the team and the process. If you end up automating, your core issue may not be addressed.
Question 3 – Let’s say I am happy with today’s process. What is it that I want to achieve by automating?
Your answer to this question depends on your definition of efficiency. If you think too many resources are blocked for tactical price decisions, and too much time is wasted not because of the complexity but because of overload, yes automation is definitely your way forward.
Question 4 – What is the level of automation that is optimal for me?
100% automation would mean no human interference at all in pricing decision. It may work for some and for some it could be lethal. Moreover for business-to-business enterprises this will surely not work, as there would be a lot of customers who want face-to-face negotiations. It’s face-to-face and not ‘screen-to-face.’
The next in line automation level would be the “operator model.” This is when a single operator or a team of operators determines the pricing rules and strategies periodically, and the machine executes accordingly. Here again the issue could be the lack of face-time.
What may be the most convenient level would be ‘automation with escalation.’ Here the machine owns the tactical process and when there is dissatisfaction with the mechanical decision (or the machine is unable to decide due to lack of inputs), there is scope for human intervention.
Successful automation, whether it is pricing or any other field, has yielded great results in the past and, yes, with advancements in technology we will see more of it in the times to come. However like any other business strategy, a decision to automate needs thoughtful evaluation. The first step for the evaluation is primitive: ask why. I hope the above questionnaire will be useful for the first level of evaluation for companies and individuals planning to automate pricing in 2016. Wish everyone a happy new year!