Myth: dynamic pricing is a black box
Danilo Zatta, PhD, MBA
Helping companies grow revenues and profits I TopLine, Pricing & Revenue Models Advisor
Dear Friends,
stating that dynamic pricing is a black box is a myth. We will deep dive this below shedding light in the B2C and B2B context.
Dave Jackson is one of the Net Revenue Management experts I admire most, successfully bringing forward topline excellence for a dozen of year at one of the world's leading CG brands, i.e. Ferrero.
He shared a feedback on the book The 10 Rules of Highly Effective pricing: ‘It's very well considered, the biggest compliment I can pay it is it considers pricing as a discipline and an art, the chapter on leadership sponsorship is so true. Can't wait to read on. As soon as I finish it I will be passing it on and I intend to get copies for my team. That way I can encourage them to read and learn together.’ Here the photo he shared:?
Please continue sharing your feedbacks.
In case you missed them, here selected posts:
Join me in Berlin, on November 20th: I will welcome you to a 1 day workshop at the Professional Pricing Society conference in Berlin on the 10 Rules of Highly Effective Pricing. Find a deep dive here.
Great pricing masterclass in Munich: join legendary Verne Harnish and me for an inspiring session about compensation and pricing to boost value, profits and talent. Join here.
Pricing Decoded: new book release revealing how some of the world's best companies took pricing to the next level. Find more details and preorder here.
Insightful pricing workshop in San Marino: join me for an insightful pricing deep dive on a weekend in beautiful San Marino with around 300 entrepreneurs. Join here.
Join one of Europe’s fastest growing consulting companies: at Valcon we are hiring. Shape the future of AI and AI-based pricing, with us. Apply here.
Myth: dynamic pricing is a black box
Dynamic pricing is fundamental for businesses regardless of the sector in which they operate, and its significance is destined to grow. If it was originally applied in the airline industry, to balance supply and demand, it has become a great opportunity for B2C and B2B companies.
Dynamic pricing enables businesses to set flexible prices for products or services based on current market demands or on specific customers. Despite the flexibility it offers, it is met with scepticism, especially from B2B companies, that consider a disincentive, or a process that requires special technology for its application. We will debunk this myth.
The mechanisms behind dynamic pricing are indeed governed by a system of rules, formulae, and algorithms. However, to consider the tool as an impenetrable black box managed by a computer means to miss the opportunities that a dynamic management of prices can offer. While there may be some metaphysical, sci-fi elements to it – cue images of gigantic machines that manoeuvre society unbeknownst to humans - , we must come to terms with the principles of this technology, especially the aspects that relate to AI, Artificial Intelligence.
It is not a question of choosing whether the setting of prices should be left to a computer or to a human being, but rather of finding a solution that combines the strengths of both artificial and human intelligence, whilst not forgetting that even the most advanced quantum computers are programmed using human language, and not the machine language magnified by tech experts.
It is the human brain that establishes the rules and parameters to feed into the dynamic pricing engine. The engine learns, elaborates, and churns out results, but the responsibility for evaluating them and making decisions about pricing level rests solely with people.?
Therefore, those who use AI tools should have some level of understanding of their mechanisms, they must be able to read the engine’s white box - where the adjective white emphases its transparency, legibility: when the calculation criteria and the parameters to be used are clear, the results are appropriate and can either be adopted or tweaked by senior managers if the circumstances require.
A positive factor of dynamic pricing is that it entails automation whilst also needing user (that is, human) engagement in the planning process, proving that it can be a useful and easy to understand tool, an added bonus that augments the power of the human brain alone by generating supplementary data, and not an impenetrable or senseless black box.
A word of warning: dynamic pricing is not miraculously going to turn salespeople into data scientists. However, if the teams in charge of setting prices are able to provide a general explanation of how the dynamic pricing engine works to the teams in charge of sales, possibly with practical demonstrations, the logic behind prices thus set will make sense to all.
A second myth to debunk is that dynamic pricing is only applicable in a B2C context, that is to say with private customers.?
I have personally introduced dynamic pricing to a number of B2B companies, with very positive outcomes, and a substantial and sustainable increase in profitability.
I remember the case of a B2B outfit selling spare parts, whose EBIT margin shot up by 180 basis points thanks to the positive synergy realised between people and machines. Technology provided an invaluable support to the business, helping it deal with hundreds of thousands of products being sold to tens of thousands of customers worldwide. To help us set and constantly update billions of single price points, we implemented an algorithm based on a series of historical data sourced internally and externally – e.g. customer types, transactions etc- as well as data on competitors’ prices mined using webcrawlers.
Through machine learning we calculated the specific willingness to pay each individual customer and product. Then we invited each sales team to review the draft proposed prices and suggest adjustments for specific situations or specific customers. The feedback was fed back into the machine, which learnt the required recalibration criteria.
The constant interaction between users, sales teams and machine has led to significant improvements: an increase in net prices and in profitability with no loss of volume of sales.
The third myth to debunk is that dynamic pricing is a lengthy and convoluted process that requires complete sets of data to be considered and implemented.? In over twenty years of experience as a management advisor, I can state that it would probably be easier to find the proverbial needle in the haystack, or the Graal, or the philosopher’s stone, than to find a company with impeccable data.
No company adopts dynamic pricing having perfect records or infrastructures. If they did, they wouldn’t need anything more than they already have!
Imperfect data sets should not discourage a business from considering dynamic pricing, nor make its implementation difficult until data are in order. Quite the opposite, in fact: dynamic pricing eventually improves a company’s ability to collect and analyse data at granular level.
What we did in many companies was start with a limited number of use cases, observed manageable sets of life cycles, and identified exactly the type of data that the customer needed.? Use cases give the possibility to introduce external data in the engine, for example data linked to the per capita income by region, or data on seasonal/temperature variations.
Dynamic pricing is not solely a technical, tactical, and data-driven tool, but is marks a process of evolution and transformation in the ethos of a company, affecting its internal processes and the perception of its value from outside.
It took a large distributor of building materials less than 8 months to move from manual price setting to a dynamic pricing engine that also returned real time information on competitors’ prices and promotions. The result was an increase in short term profits by 52 million euros and a concomitant growth in volumes of sales –12% in the first year alone, with no attendant negative impact on market or price positioning.
B2B businesses that successfully implemented dynamic pricing started off from a single use case? ?and progressively expanded until - using a reasonable mix of human wisdom and technological capability - they found the perfect format: prices that were sound but also could be modified at relevant times and personalised by customer.
What is your view on this myth?
Interested in learning more about pricing myths? You will find pricing insights in the book The 10 Rules of Highly Effective Pricing.
After the best seller The Pricing Model Revolution, this new book by Dan Zatta deep dives new key topline topics, proposing to CxOs simple but effective rules to lead the Pricing Transformation. With the support of several concrete cases Dan also shows how in this journey C-level pricing attention is a key factor of success to reach above average profitability.
Luigi Colavolpe, General Manager & CFO UniCredit International Bank (Luxembourg)
Get your copy of ‘The 10 Rules of Highly Effective Pricing’ here.
Get your copy of ‘The Pricing Model Revolution’ here.
You are most welcome to share your views, feedbacks and own pricing experiences. Thanks a lot for your interest and support!
?
?
Strategy | People Leadership | Problem-Solving | Results Driving | Innovative | Curious | Adaptable
1 周?? "dynamic pricing is not miraculously going to turn people into data scientists," reminds me of when I was rolling out first gen retail price optimization software. I had quite the time convincing some users that science did not mean magic. ?? There was a literal perception you could just tell the software to come up with prices that would increase sales (a lot) and profit even while being uncompetitive in the marketplace.
Strategy | People Leadership | Problem-Solving | Results Driving | Innovative | Curious | Adaptable
1 周At the risk of muddying the waters, dynamic pricing is probably too associated with computer generated retail setting. Especially in the context of electronic shelf labels, dynamic pricing is simply "frequent" (a very subjective term) price changes. When cast as a villain, the frequency of changes is implicitly due to swings in demand. But ultimately dynamic is simply benign change. However model driven price setting ("popularized" with models doing surge pricing) frequently is highly opaque. In fact early B2C price optimization engines were widely rejected by users for being "black boxes" which to the user meant neither the software nor its purveyors could explain the why behind a retail. Many software companies recognized that the lack of transparency hindered adoption (and sales!). So they responded to customers and provided greater visibility to why an engine suggested a particular retail. Thus computer set retails are not by themselves "bad" either. Black boxes though are unhelpful. A human setting a retail but not being able to explain why would not last very long in a pricing job. It is appropriate that machines be held to the same standard. Pricing Art = (good) judgement + tools (for efficient automation at scale)
AI Founder & CEO @ FutureUP | Building the Future of Price Optimization | Top 50 Thought Leader in AI | Raised $9m in VC funding in AI
1 周One of your best newsletters, Danilo, addressing so many myths and challenges!?????? Let's start with the first: Is Dynamic Pricing a black box? Well, it shouldn't be, but that depends on the AI model used. It would be if it is based on a general-purpose AI like Generative AI. This is why it's better to use AI models with econometric/business intelligence, which are explainable & more transparent. Re the applicability of Dynamic Pricing to B2B vs. B2C: Indeed, it can be applied to B2B but in a different form than B2C, eg, not in real-time. But it has more difficulties than B2C, and you can find some here: https://www.futureup.io/post/navigating-ai-driven-pricing-challenges-in-b2b Re data, no company has ideal information. B2B has no big but “small” data. A lot is proprietary, like competitive information, and not available. But AI can predict and, fill in missing info to some extent. This can be tested to see if the model works well with less data. If not, you can add data proxies or perform market/pricing research until you get the desired accuracy. Using AI models with econometric intelligence is critical here to ensure accuracy with less data.
Business Solutions & Digital Transformation Consultant | AI Technology
1 周Great post! I agree;?finding a solution that combines the strengths of both artificial and human intelligence! This will be a game changer
Helping customers implement the future
2 周I would expect that dynamic pricing in the B2B world will become more common as B2B buying journeys and behaviour begin to imitate the B2C journey more closely... Great post as always Dan!