Companies across virtually all industries are looking for ways to drive profitable growth. Over the years, they have invested in technology to streamline and automate a variety of business processes. This includes implementing Enterprise Resource Planning (ERP), Customer Relationship Management (CRM), Demand Planning, Inventory Management, Transportation Management, and Warehouse Management applications. After capturing many of the “low hanging fruit” benefits of these solutions, companies are now asking, “Where is the next big opportunity to realize significant business benefits?”
For many companies, the answer is Price Optimization.
Price Optimization is not a new discipline or software category. Back in 1992, for example, Michael V. Marn and Robert L. Rosiello discussed the link between price management and profits in their seminal Harvard Business Review article, “Managing Price, Gaining Profit.” According to Marn and Rosiello, which they illustrate with various case studies:
The fastest and most effective way for a company to realize its maximum profit is to get its pricing right. The right price can boost profit faster than increasing volume will; the wrong price can shrink it just as quickly…For companies with average economics, improving unit volume by 1% yields a 3.3% increase in operating profit. But a 1% improvement in price, assuming no loss of volume, increases operating profit by 11%. Improvements in price typically have three to four times the effect on profitability as proportionate increases in volume.
Over the years, various other studies have supported these findings. Alex Abdelnour and Walter Baker from McKinsey & Company, for example, highlight in a September 2019 article how “pricing can do more than traditional margin-expansion methods to create new value” for distributors. According to the authors, “An average distributor in 2018 would have to grow volume by 5.9 percent while holding operating expenses flat to achieve the same impact as a 1 percent price increase — no small feat, especially in mature markets where competition is fierce and growth often comes at the expense of profitability.”
Companies are also looking for ways to leverage Artificial Intelligence technology to improve their operations and achieve business benefits, but to do so without having to rip out and replace their existing systems and processes, which would be very expensive, time consuming, and disruptive.
In a recent Talking Logistics episode, I discussed these two topics — price optimization and AI enablement — with Alessandro Chiaramitara, President at Delly’s, and Michael Romeri, CEO at Analytics2Go.
If you’re not familiar with Delly’s, the company is the Brazilian leader in food service distribution, similar to Sysco in the United States, with about $1 billion in annual revenue. The company serves over 200,000 customers, including restaurants, bakeries, hotels, pastry shops, and supermarkets, offering them a large assortment of products (more than 25,000 SKUs). Delly’s has a large distribution fleet and more than 20 distribution centers and cross-dock facilities across the country, enabling it to serve 70% of the country.
In short, as Alessandro put it, “it’s a very intense and complex operation.”
Why Focus on Pricing for AI Enablement?
I began the conversation by asking Alessandro, “With regards to AI enablement, why focus on pricing first?”
“As a business leader, I believe you always have to think about your priorities and what will change the landscape and have a big impact on the business,” explained Alessandro. “When you talk about food distribution, the pricing aspect is very relevant because it has a big impact on our bottom line.”
Alessandro highlighted several factors that complicate pricing at Delly’s:
- Volatility in commodity prices: “We deal with products like beef, chicken, rice, soybeans, and other commodities that are very exposed to [frequent] price changes in the market. If you don’t react properly in terms of pricing, you can lose volume or reduce margins.”
- They work with a lot of “Mom & Pop” customers: “We do not have many large chain customers with annual fixed contracts. For most of our customers, we have to change our pricing every week depending on our acquisition costs, especially for commodity items.”
- Pricing involves both Category Managers and Sales: “We have two big two important players involved in pricing: Category Managers, who buy the products and know about the trends in the industry, and Sales people, who know our customers. So, to come up with pricing, it involves combining the know-how of these two groups in order to maximize profitability, and these two groups [sometimes have conflicting interests].”
So, back to the question, why focus on pricing for AI enablement? Alessandro summarized the reasons why in the short clip below (spoiler alert: the large volume of data involved is a big part of it):
As an example of the data that must be combined and analyzed to dynamically change prices, Alessandro highlights what it takes to price salmon in this short clip:
When you take all the SKUs and customers they have into consideration, “we generate more than 800 million data points per month,” says Alessandro. There is no spreadsheet big enough or human brain powerful enough to do such an analysis!
Challenges with AI Enablement
I asked Mike about AI enablement and the challenges companies typically face doing it.
“You would think that supply chain would be the perfect choice for AI because AI works best where the complexity or the rate of change of data is great, and as a result it’s very hard for human intelligence to deal with it in an effective way,” says Mike. “However, there are three main challenges companies face with AI enablement and they’re interrelated.”
One big challenge is data orchestration, as he explains in the short clip below:
“Getting the data right and making the data work is still the longest pole in the tent on almost all our projects,” says Mike.
Another challenge, not surprising, is change management.
“Change is hard for all organizations,” says Mike. “So, if you redefine the process, it means you have to retrain the people. In Delly’s case, they had about 2,000 salespeople at the time we first implemented, and with Alessandro’s guidance, we decided not to change the interface for any of those sales people, so we switched from human intelligence-based pricing to AI-based pricing over a weekend and they started getting new prices but it was in the same format and the same request process they were used to.”
I asked Alessandro about the benefits they have achieved to date since implementing the Analytics2Go solution and AI enabling their pricing process.
“We have improved our gross margin by 1.1 percentage points, which for a company of $1 billion dollars of turnover is huge, and we have reinvested that money back into the business to accelerate our growth,” says Alessandro. “Another very important output is that we moved from price lists by state to customer-specific price lists, which has had a huge impact too.”
For more details on this change in price list approach, which includes customer clustering, please watch the full episode where Alessandro and Mike share their insights and advice on many other questions, including:
- What were some key capabilities or requirements Delly’s was looking for when evaluating technology solutions?
- Why did they ultimately choose Analytics2Go as their partner?
- How did they go about implementing the solution?
- What advice would they give to companies that are just getting started on their AI enablement journey?
After watching, post a comment and share your perspective on price optimization and AI enablement!