The AI-driven Crunch Markdown Assistant makes it easy for fashion retail teams to set the right end-of-season discounts or price markdowns, maximize revenue, and clear out stocks just in time.
Deciding on appropriate end-of-season discounts or so-called price markdowns is notoriously hard, takes time, and therefore often leads to disappointing results. Crunch Analytics' Markdown Optimization Assistant ensures no money is left on the table, using the latest advancements in big data & AI-techniques to perform the heavy lifting.
end-of-season sales
lead to dissapointing results
NOT KNOWING HOW THE SALE OF AN ITEM RESPONDS TO A CHANGE IN PRICE MARKDOWN
HAVING AN UNCLEAR OR INCORRECT DISCOUNT OBJECTIVE
LACKING A USER-FRIENDLY TOOL THAT ENABLES A SEAMLESS PROCESS
Our markdown optimization assistant suggests the right price markdown for any item in your catalog. An optimal discount that should sell the item out just in time, while generating more revenue and a higher margin. By automating manual tasks and taking over the heavy lifting, the solution not only enables better results but also cuts the time spent on the task by half.
“We were able to achieve a +30% revenue & +8% margin increase, applying the price markdowns suggested by the solution”
In this video, Luc De Baets - Buying Director at Torfs - explains why Torfs chose to adopt the markdown optimization solution and how the algorithm-driven toolkit helped the team to achieve astonishing results.
Put an end to those endless spreadsheets and the manual gathering of data. Obtain a clear view on price elasticity and predicted impact and get the job done with a few clicks of a button.
Automatically gather data from a variety of sources and send price markdowns to your ERP system with the mere click of a button.
See the predicted impact of a price markdown and understand how & when discounts can help clear out stock that threatens to pile up.
Have a solution calculate the optimal price markdown taking into account sales figures, inventory levels, price elasticity, timing, residual value, etc.
Easily review relations between items in categories, channels & countries and adjust where desired.
Receive price markdown suggestions that are tailored to your unique organization and which take into account constraints such as your pricing strategy, store-related limitations, etc.
Use the management & tactical dashboard to ensure your teams can apply their wisdom & experience to review markdown suggestions before they take effect.
Import data from
various sources
Validate price
markdown suggestions
Send to
ERP tool
Every retail or e-commerce operation has a unique situation in which it operates. Be it business and operational constraints or specific strategic rules. We sit down with the team and ensure the solution accommodates them all.
We identify all relevant sources of data and ensure an automated and reliable transfer of data can occur. Such data may include product information, historical transactions, rotation speeds, inventory levels and many more.
Once a reliable transfer of data from all relevant sources is available, we train the applied algorithm. Using historical data, we are able to uncover patterns and indicate a theoretical yield. As we sift that information through various business and operational filters and match it with more recent data, price markdown suggestions emerge.
In this step we prove the algorithm works in practice by preparing an A/B test design. During the next sales season, the algorithm will set prices for a subset of items, while your team uses its traditional methods for another subset of items.
With a successful pilot under our belt that proves the business case for the solution, we bolt the toolkit down and ensure its robustness. The solution can then become an integral part of the retailers pricing process, ready to be used during upcoming sales seasons.
The solution can be further customized, adapting it to additional constraints, new strategies, or other factors that have an impact on the price markdown process.
Not confident about what actions to take next, using big data & AI-techniques in your retail pricing processes? Would you like to learn more about the opportunities these technologies offer and how others are using them? Let us know
How a custom-made algorithm reduced a client’s time on managing inventory levels.
Leading retailers all over the world are upgrading their customer experience through data- and AI-driven marketing. How does a Belgian retailer compete with these online, heavily personalized retail giants?
A price markdown or retail markdown is the permanent reduction of the selling price of an item, to anticipate the end of its presence in a retailer’s seasonal or overall assortment. It is typically used to clear out inventory or stock by the end of a season or the end of its life cycle.
Contrary to regular discounts or promotions which are often ‘temporary discounts’ or aimed at a specific customer segment, price markdowns are ‘permanent discounts’. They are used at the end of a season - or the end of a product life cycle - to increase sales velocity and sell out the item by a given end date.
In some retail & e-commerce sub-sectors such as fashion, price markdowns are essential as the assortment is seasonal & trend-sensitive, and the residual value of a given item can be a mere fraction of what it was at the peak of the season. In customer electronics, on the other hand, there is the pressure of technological advancements that impacts the life cycle of a product.
Markdown optimization is the process of setting the appropriate markdown at a given moment in time. Every retailer needs to balance the impact of a price markdown on the sale of that item.
If a price markdown is too high, the item might start to fly off the racks. Such creates turnover but might clear out stocks before the end of the season. And it probably cuts too deep in margins. The retailer is confronted with empty shelves or has to mark the item on its webshop as ‘out-of-stock’, possibly losing clients to another retailer.
If a price markdown is too low, the stock clearance may take too long and leave a retailer with left-over inventory that has lost a large portion of its value. The loss in turnover might hamper the retailer's ability to purchase enough inventory for the next season to come.
Our markdown optimization solution, therefore, suggests what the most optimal price markdown is, week over week. Gathering a tremendous amount of data on all items, it applies machine learning algorithms to predict how customer demand will respond to a given markdown. As we sift that information through various business and operational filters and match it with more recent data, price markdown suggestions emerge.
Every retail or e-commerce operation has a unique situation in which it operates. Be it business and operational constraints or specific strategic rules. That is why we deem it important that retailers don’t purchase off-the-shelf software.
The most important aspect of the challenge isn’t working with these brand-new, promising technologies. It is making sure you target them right, making sure you aim for the appropriate optimization objective. For that, you need the combination of an advisor with both technical & sector-specific business expertise.
For retailers taking the step to introduce a new piece of software that is able to partially or even entirely take over tasks, it is important to get the full team on board.
In the initial stages of a project, we ensure that the objectives of both the business and technical stakeholders are aligned. Throughout the implementation of the solution, we work in ‘sprints’, to ensure that everyone stays aligned.
Moreover, we kickstart each project with a pilot to demonstrate the business case. During the very first sales season, we introduce an experiment, an A/B test. We divide a set of items into two subsets. One subset will be priced, using the traditional methods applied by the team. The other subset will be priced using the markdown suggestions provided by the algorithm. By the end of the season, we compare results and tweak where necessary.
In the first phase, we provide a set of services to get a retailer up and running. We connect to the client's data sources, find key patterns in terms of price elasticity based on previous sales periods, and train the markdown pricing algorithm accordingly. An algorithm that is further refined during the actual sales season, using the input of more recent information.
Once the sales season starts, we enter the second phase and a week-over-week sequence kicks off. Last week's sales information is downloaded and validated. Based on several parameters, the solution calculates new optimal price markdowns for the upcoming week.
This and additional information is displayed in two dashboards that enable follow-up and should instill trust in the process.
The first is a 'management dashboard' giving the category or pricing manager, a general overview of the situation. How are we doing in general? How much are we discounting? How are they distributed across different product categories? What are the trends? How are clients responding?
The second dashboard is a more 'tactical dashboard' showing information on the level of an individual product such as optimal price markdown, how the product has performed, how discounts affected sales, etc. It allows the category/pricing manager to take into account the suggested price markdown for a given product and evaluate it. The dashboard also indicates priorities, helping that same person to focus on the products where the potential for additional revenue is the greatest. If all is validated, the output can be easily uploaded to the client's ERP- system.
The best part of it all is that this whole sequence can be partly/fully automated, making the category/pricing manager an observer of the process, able to spend more time focusing on strategic analysis and discussion of results.
Each situation is different, as retailers currently work with a variety of systems in a variety of environments. Yet, we are convinced that we can get most projects up and running in less than four weeks' time.
