How to leverage AI-powered assortment planning in 2024
For every fashion retail team, no time of year is without its challenges.
All year round, brands and retailers are trading over 3 timelines at once: current season performance, past season analysis, and future season planning.
No matter your role, your brain is always across three zones — and the challenge we face is the complexity of managing all of that simultaneously.
Multi-faceted assortment planning
When you buy any given product, just for that one single product you automatically have many crucial decisions to make:
What are you going to price it at? How many units? The locations, the size runs, the margin targets… and that’s just the high-level decisions.
Your task then, is to perfectly balance that assortment. Wherever you are on the globe, you want to make sure that you’re sending the right seasonal product to the right place, and in the right category mix for the different consumers.
Then, you must consider the colors, the price bands, the silhouettes, etc.
Yes — it’s a lot — which is why teams are looking to simplify retail’s biggest challenges with advanced technology.
Key Buying and Merchandising issues facing fashion teams
For buying and merchandising teams the challenge has always been the level of data that you could plan at, but can’t because you're being slowed down with BI tools, spreadsheets and pivot tables.
The struggle is not being able to keep up with the multitude of changes: dates, attributes, fabrics, and delivery schedules with the need to consolidate sheer volumes of data — and just not having the tools to do it.
The biggest challenges facing fashion teams working without advanced system processes are:
- Buying the right product, at the right level of data integrity, and not be left with excess inventory
- Not leaving revenue on the table when best-sellers are selling out faster than expected
- Creating a ‘no dead end’ experience for the customer journey when understocked
AI: Understanding its value
For all teams, it’s clear data points are expanding exponentially with every new plug-in and platform.
So, how do you make sense of all that data?
Artificial intelligence should be used to solve a lot of the heavy lifting of data consolidation, analysis and decision-making.
Right now, teams and businesses see the value of AI on a scale of 0-10, with 0 being complete hype and not useful for my day-to-day, and 10 being deeply valuable and role-enhancing.
Most businesses sit around the middle mark of 5, with the issue being they are unsure how to test for value.
The key is for teams to use AI processes that are practical and auditable.
To measure the value of AI platforms and tools, businesses can come to their own conclusions by asking and discerning:
- How much added revenue is this making me?
- How much time and cost is this saving?
- Can I interrogate and understand the logic it’s using so I can trust it?
Combining AI deep tagging and data analytics
Deep tagging capabilities offer insights that go beyond being able to see what’s been selling from specific tags and attributes.
It can deploy highly targeted and rich product attribution at a scale and speed that is not humanly possible.
This kind of dynamic product tagging trawls deeply into historical and current style data and imagery, identifying a unique and enhanced set of style attributes for every product — hours of work crunched in seconds.
The advantage of Style Arcade’s deep tagging for example, is the ability to consolidate all of these attributes and detect overall demand opportunities you otherwise would not have had access to — and apply these learnings to adjust your purchasing decisions to meet this demand.
Use cases
- When it comes to future assortment planning and analysis, you can filter your products right down to their performance based on geo-location (even weather), price points and acquisition. How your short-sleeved v-neck t-shirts perform in your LA stores compared to your New York stores matters. With completely different consumers in the same country, you must be able to see and keep up with demand.
- Additionally for marketing and online promotions, deep tagging allows a new level of personalization and segmentation. You can create emails with product feeds that consider their geolocation preferences and the stock level at the time the person opens the email. This way, you’ll never show customers something they can’t convert on.
- Unifying the taxonomy (dictionary, tagging lexicon) on the front and back ends, will help with enhanced search; expand the filters you have in your navigation and help your customers locate items easier. This will also boost SEO - the holy grail of attracting new customers today (e.g. does your taxonomy dictate, ‘leopard skin’ or ‘leopard print’). As well as help you seamlessly buy to the deeper attribution demands.
- Buying and merchandising teams can automate new buy and size quantities based on how products with similar product attributes performed, and range new products based on rich historical performance data.
AI: Generative AI solutions
Generative AI has paved the way for retailers to start to look at real-life applications, and we’re seeing that not only on the technology side but on the B2B side and for the customers. This is beginning to take shape in the following ways:
- Product descriptions - leveraging Gen AI and AI tags to automatically generate product descriptions and unique titles, and edition manually for full control over your website content
- Styling enhancements - from ‘shop the look’ for on body images to ‘shop the style’ for standalone products in order to mimic styling in the most live way possible
- Smart collections - leveraging styles, occasions and trends in order to build recommendations for increasing the basket size
- Customer service - allowing teams to recommend similar products in a customer’s size by using the search similar functionalities, or letting them know the exact date their products are expected to be replenished because it's connected to all back-end systems. You can also enhance intelligent chatbots, virtual assistants, and self-service to address advanced consumer inquiries
- Social content - Recognize patterns and trends in viral content and create new content based on your recommendations for your brand. And, prompt video platforms to create short-form videos for social media platforms
Originality and creativity are still key to your content. When you get the partnership between your brand’s creative vision and technological proficiency right, you can continually maintain a competitive edge.
Summary
- AI in retail and e-commerce is booming, but not every solution will provide the performance, revenue and cost-saving boost brands expect to magically occur
- Testing and having clear ROI measurement is key
- Tackle assortment challenges both on the front end and the back end
- Utilizing data for forecasting is crucial to meet customer demand and avoid overstocking
- Assortment planning is easier with visual AI, discover the deeper level of what is selling and in demand and plan your inventory investment accordingly
Want more in-depth knowledge about the opportunities of Ai-powered assortment planning?
Watch the webinar here.
Image credit: Rachel Koukal
