There Aren’t Necessarily ‘Good’ or ‘Bad’ Promotions, Only Good or Bad Promotion Combinations

Planning effective and relevant promotions is an ongoing exercise in adaptability. There are complex and evolving factors at play that impact product availability and consumer behavior. Right now, retailers are running into challenges with executing planned promotions because of supply constraints and inflation-induced margin squeezes. Retailers don’t want to promote items they may not have the stock to support, and rising costs mean they may be unable to hit promotional price points that worked well historically.

Because of this, retailers are focused on prioritizing promotions that will achieve business goals, ensuring that investments will yield desired traffic and sales. But that’s easier said than done.

Promotion planning is like chess, where you must think several steps beyond your next move. Without perspective to contemplate the possible what-ifs and strategize accordingly, the human mind tries to make it very simple: is this decision good or bad? But nothing is necessarily good or bad by its own nature, especially in business. It’s about interdependencies with everything else happening around it, a result of factors both in and out of your control.

Because the success of promotions depends on an interconnected network of decisions, it’s important to know what planners are up against today, and what steps they can take to be most impactful in their roles.


Promotional Planning can Feel Like a Series of All-Consuming Micro-Decisions

Anecdotally, we haven’t seen a significant decrease in trade promotions funding from vendors. There was an obvious downturn early in the pandemic and month-to-month seasonality beyond that, but promotions are still incredibly important.

Even so, conversations about promotions are happening on an ongoing basis between retailers and manufacturers, and even with annual and quarterly plans established in advance, things can change through the year, often on a weekly basis. Planners struggle with uncertainty, asking “Is this the right decision to make?” After all, earning their bonuses depends on getting it right, so they have a vested interest in hitting those performance targets for the business.

To overcome complexities, many retailers have applied basic analytics to promotion planning, but planners still don’t feel they’re doing a good enough job, even armed with accurate and real-time data. There are multiple dimensions to consider.

  • Is this the right product to put on promotion?
  • Is it the right time to promote this item, or should I wait a week?
  • Is this promotion driving households and/or incremental sales?
  • Is this the right offer to make in this promotion, and am I using the right vehicles?
  • Should I cancel a promotion, or reduce its frequency?

Each of these factors interacts with the others. For a human being to make sense of all the variables, determine what role each factor plays and identify what factors are working symbiotically in the same direction (versus working against once another) is fundamentally impossible. That’s why retail promotions are quite amenable to the use of artificial intelligence for decision making.

Optimal Promotion Planning Leverages AI to ‘Understand’ it All

Because any number of factors impact sales, revenue and profitability, retailers need AI to bring a sense of comprehension to the possibilities. Consider what-if scenario analysis — the human mind can easily weigh outcomes of two, three, maybe five promotional scenarios. It’s impossible for any individual or team to do 20 or more what-ifs and understand how each one impacts another.

Across time, across products and across regions, AI can reveal where a promotion cannibalizes sales of another high-margin item, for example, or where a promotion isn’t performing because of too many promotions in a category. AI supports trade funding redistribution broadly, in addition to more targeted decisions associated with customer-insights-fueled personalization in marketing.

AI turns the typical planning exercise “I’m going to do X; what will happen?” into a different conversation altogether. The planner says, “Here’s the outcome I want,” and the AI responds, “Here are the recommended actions to get there.”

That’s not to say human intelligence goes out the window. The system may return a recommendation that doesn’t make business sense. Soup isn’t best promoted in the summer, for instance. Humans need to set these guardrails and specify business rules. Another rule could be never having a premium product priced lower than a key value item. At the end of the day, the use of AI will build user confidence, and make planners better at their jobs.

Back to the chess analogy. When applied to promotions, AI doesn’t know a good move from a bad one — it just knows the path to a victory, and only the minimum number of actions it will take to get there. There’s no ego, no superstition, no preconception of what someone is ‘supposed’ to do or what has delivered results previously when market conditions and shopper preferences were significantly different. Victory in retail is about the right combination of promotions — a full promotion calendar — that, at the end of the day, generates the target financial performance.

Rahul Bhattacharya serves as Chief Analytics Officer, Analytics for Symphony RetailAI. He is an Applied Data Science leader providing technical consulting and leadership for analytical products, solutions and projects for customers worldwide. He architects and manages the implementation of business applications using advanced analytics built on state-of-the-art predictive analytics tools, technologies and methodologies.

Feature Your Byline

Submit an Executive ViewPoints.

Featured Event

Join the retail community as we come together for three days of strategic sessions, meaningful off-site networking events and interactive learning experiences.


Access The Media Kit


Access Our Editorial Calendar

If you are downloading this on behalf of a client, please provide the company name and website information below: