Episode #0102 - Can a pricing IT tool fix B2B pricing

Pricing College Podcast

In today’s episode, we discussed what are the pitfalls of seeking a pricing tool to run B2B pricing.

Notes on the time-stamped show:

[00:00] Introduction

[01:06] Joanna argues that without a great pricing framework and architecture, businesses cannot expect optimal outcomes from automation.

[03:31] The field of pricing have two approaches to technology. The first is revenue management systems, and the second is optimisation systems.

[07:13] What processes do you need to set up in order to make the most out of these tools?

[12:03] Aodhan talks about how important building a good value management system is before businesses can employ the appropriate computer systems.

[13:58] Category management and pricing teams should work together to properly quantify value.

At Taylor Wells, one question we get asked quite frequently, I suppose because we focus on the B2B and the B2C sectors, is what computer system, which IT system, which new fangled new technological approach will do the job for us, will really encapsulate our pricing strategy, and what should we implement. To some extent, the answer is not often what people want to hear. People, I think noticed in 2022, believe that machines can and should do most things for us, we're used to typing and google and then coming up with the answer. But, I think, when it comes to B2B, B2C pricing, tools have a real role but they will not replace the human touch.

Yeah. To put it simply, I think a lot of pricing systems they're great, if you got a great framework and architecture in place already, then you can automate that. But often, what they do is automate what you've got so if you look at it in the negative, you've got broken poor systems, you've got no price structure, you're discount levels are incorrect or you don't have any, you've got discretionary pricing, there are no price controls. Then really what's the point of getting a high powered pricing system to automate that, what you're just going to get is raw automated junk in the machine calculating incorrect and often cost-plus pricing very quickly. So, in a way, what happens next, what people do often is well they stood by that there is a system, a silver bullet to correct what is fundamentally a broken architecture. And often if you got a broken architecture, it's misaligned with a business module and operations. This and in a way indicates that there are some business strategy changes and operational changes that need to occur as well. But regardless, what happens is that maybe a senior executive, the CEO buys this new pricing system hoping that it is the silver bullet to correct everything, may misunderstand the initial sales pitch from the vendor of that machine. What happens then is the vendor comes in, plugs it together, they call it integration with your other systems, like your ERP. And they find that, yes, lo and behold the pricing architecture is broken too. So they work with the business strategy trying to correct that. But often, that leads to a very long drawn up process and very costly process for the business as these vendors are very expensive and end up staying there for many years and not really fixing the actual problem, and just automating it, fundamentally. Aodhan, what do you think?

I love the trends. I read an article once that humanity has not really moved on since the 1950s, nearly all the technologies that we have were existing in some format at that point. You know, jet airlines, motor cars, all that sort of stuff, antibiotics to a large extent. And all we've had really is computers and electronics in the last 20/30 years which have grown infinitely more powerful than they were even in the mid-80s. But the negative of this is that we've become so focused on big data, data analytics, statistical analysis, and the big data that the internet has given rise to. So if we look at the pricing world, we have two real approaches to technology in that aspect--in computer programs, we have really the revenue managements systems which are implemented in airlines and capacity-constrained businesses, such as hotels, tourism, cars. We've seen them try to be implemented in tool hiring less successfully. And then on the other end, you have what I would regard as growth from A/B testing, almost like a website optimisation system based on pricing such as Price Intelligently. There are two things both of these have in common. They have the ability to measure people coming to something and then the historical results of what happens. So you can show them a different pricing presentation, everything else is equal. Statistically then, you can draw conclusions as to prices that will optimise sales, decrease sales, etcetera. That's in the Price Intelligently on that aspect and then on the revenue management side, you know you're selling x number of seats, historically know on a Monday, x number of people, statistically will look at this category and then you can optimise the sales with statistical variants with the risk weighting, etcetera. You can be quite confident in that. What I would say, is that some big numbers when you have statistically valid samples. But when you're in a B2B environment, you're quoting, you're doing rendering, you're probably aren't into statistically valid numbers of things. The example I'd give is, you look at an auction business, you know you're selling a painting but you're not using a revenue management system to sell it and the reason is there are no statistically valid numbers behind that. And so in B2B and B2C pricing, when there's not so long line, it becomes more difficult. You will probably see it in a civil market where there's large footfall, where people are using cards, etcetera to come into the shop. You know what they're buying, you could measure aspects in that regard. There's that grey area where there is room for these optimisation techniques certainly. But when we're looking at more, for traditional B2B, you might be only working with 5 or 6 customers, you don't really know how many people are looking at you, you're not capturing the data as to how many people have asked about your pricing. In that instance, it's extremely difficult and you just aren't capturing the information to feed it into a system to be able to really use those for there to be authorisation approach or the revenue management optimisation either.

That's true and that's what I was referring to in terms of often that's a broken pricing architecture just because it doesn't happen in B2B very often doesn't mean it shouldn't happen. I agree with you in a sense, to make the most out of these tools, you have to set up these processes, measurements, and tracking prior to buying the actual to all make it worthwhile. But often, that particular piece of work is left because businesses in B2B believe that if they just buy the system then that will correct everything else. But it doesn't. So again, I agree with you in the sense that, the pricing system is very effective at doing good pricing analysis. It calculates accurately. However, what it doesn't do and what you need to do before buying this system is set up the business rules and parameters, the conditions and the scenarios that you want to test. And then use those analytics, so set up the ratios, the measurements, the tracking tools. This is all, I call a price architecture. And this really does take two years to do. Get that piece of work done before you buy the system. And if there's one thing that you should take away from this, is that don't go to the system first because it doesn't build your architecture. It doesn't give you the learning that you think it will right away. What they will say is, you need that all set up in the first place, you need the tracking tools, you need your ratios, you need your quote to book, how much of your revenue is contracted versus uncontracted, how many of your products are specific to customers--there's one to one pricing, how much of your revenue is uncontracted, so you have many price points in customers. Because then you'll have different ratios, and different trackings, so you'll know how to optimise different types of revenue groups. If you've got those answers and those things set up, yes automate but don't do it before because you really won't get the answers, just gobbled nonsense.

"If you can't measure you can't approve it." It's a famous mantra from some management gurus. But what I'd say is the closer your business is to commoditisation, the more likely you can capture statistically valid information, measurements, quotes to book, all those metrics that we discussed. You know when you're setting large numbers of products, this is just my viewpoint but when you get into more bespoke stuff, when you're probably dealing with fewer customers, potentially you have fewer competitors in the market, you're value adds or maybe less more to your business, whatever they could be. I personally think that the opportunity for the value of a good sales team in that instance, a good marketing team, a good pricing team, and the human element is more important. Even if you capture all that information, you go through that process, the information you capture in the past, if you're business is constantly evolving, constantly delivering new stuff, the product you give this year different to what you give last year. If the market has changed, and your product has improved, is the information from last year statistically valid? If we're talking about revenue management and the airline, you know flight into Chicago, from New York, for 9 o'clock on a Monday, excluding Covid of course, clearly, there are historical precedences that make sense. But if your product is different, if it has really changed, if it's new, in those instances, the statistical aspects offered decrease. I think a lot of it will come down to your valued management system, how you articulate that to your customers, and your ability to build a sort of network of facts. You'll get it into real complexity, and the more complex things get, it's much harder to put them into a cookie-cutter style system. So you need to be careful. What I would almost say, if you're focusing on being very driven by a system, you should build your value management to suit the system, rather than, which is what Joanna talked about, building your computer system to suit your value management system and strategy. Because the more complex and better your value management strategy is, potentially, the less likely an all-consuming computer system will suit you. Tools are really useful in small aspects from mechanising and automating stuff that humans are probably not best suited to do, to boring, monotonous work that could be done quickly. You know quoting, emailing, CRM systems. But sometimes we can lose track of what really important here.

Yeah, it reminds me of the client I'm working on at the moment. I'm working very closely with the category team to understand at the skew level the value of their product failure, and that really for the pricing people, they are looking pretty much at the attributes of the product. That's the first step, the second step looking at the value of those attributes in the eyes of the customers. That's a different type of cognition that a computer can never really capture and when you look at pricing systems, they just stop at that statistical analysis. They don't go into this cognition that I'm talking about. That real value-based perception and willingness to pay because it just simply can't. AI learns but it doesn't learn like and I have not to date seen a system that thinks in that way. So this is the value of having a great category management team working alongside pricing cause only together can you really unlock and quantify what value is. First, you've got to define it and then the pricing manager works then quantify that. And quantifying is a testing process. You start with your hypothesis, once you've unlocked the value and you've laid it out. But then you've gotta test it in a market and you have to look at price response and actual feedback from the customer. Again, different types of feedback, not just price response sensitivity, and elasticities, we're looking at the why as well as the what. So this is why a lot of AI just, can't do that sort of stuff for B2B businesses. But there are parts of B2B businesses, you know in terms of automating quoting tools but again, a quoting tool for B2B needs to be thought through first by people to make sure that it fits in with the business strategy. Okay, I think that's all I have to say but if you have any questions for either of us please feel free to reach out. I'm more happy to talk to you about that.

Yeah, listening to this podcast today, makes me feel like a lot from the industrial revolution so this weekend I'll be heading out with a baseball bat to smash up computers and machines. Join me if you feel free. Have a great weekend.