Driving Seasonal B2B Sales Efficiency with Machine Learning
Your best customers commonly create the lion’s share of your sales, and you should do everything you can to ensure they feel loved and supported
But what about your other customers? How can they be efficiently supported without the same time investment that your top-tier customers require? The answer can be found in data science and automation.
Let’s back up to understand this dilemma. For manufacturers that sell products to retailers (B2B), sales reps can take hours to create personalized offers for the top few retailers. Typically, these top 20% of retailers represent 80% or more of the revenue and deserve this individual attention.
However, to grow efficiently, the other 80% of retail customers that don’t contribute a significant slice of revenue should not require the same time investment. To improve the efficiency of lower-priority sales activities, machine learning can be implemented to provide sales teams with comprehensive customer insights and automated personalized offers.
While there are various types of B2B sellers that would benefit from this type of technology, this blog will primarily look at how seasonal B2B sales teams could use machine-learning to improve their processes.
Seasonal B2B Sales Process Overview
Seasonal B2B sales deal with products that change every season. This includes products like apparel, sports equipment, outdoor furniture, etc. The sales process for businesses that sell these type of products typically flows like so: the retailer orders months before the season, the products are manufactured, and then the retailer receives inventory for the season.
However, before retailers order products, assortment plans (process of selecting the collection of products which will be offered in particular regions for each type of retailer) are created by the product team for each customer segment. Once the assortment plan is complete, the sales agent receives and reviews the assortment plan for each retail customer and then uses the plan to create a personalized offer for each retailer. From here, it’s the sales agent responsibility to monitor and help retailers fill their orders, even extending support online through the B2B ecommerce portal.
These seasonal B2B sales activities, from creating a personalized offer to monitoring customer’s order, require a significant time investment for small and mid-sized retail customers alike. With machine learning, however, this process can be automated to free up time that’s currently being allocated to low priority customers – ultimately, creating a more efficient and impactful sales process that drives more revenue.
Establishing Product Relationships with Machine Learning
Understanding the role machine learning can play in optimizing the B2B sales process requires knowledge about how machine learning algorithms establish relationships between products.
In particular, there are three product relationships that are critical to the insights generated by these technologies: product lineage, similar products and complementary products. From these product relationships, machine learning surfaces an automated personalized offer based on the assortment plan.
Product lineage is the linking of a previous seasons’ product to the most similar product that is for sale during the current season. These can be direct replacements with minimal changes, such as last season’s Hiker1 Backpack becoming this season’s Hiker2 Backpack, or this can be less obvious relationships, like last season’s Day Hiker Pants are a replacement for this season’s Soul Pants.
With 4-Tell’s machine-learning technology, for example, product lineage is calculated by matching the product name and product attributes from previous seasons’ products to the current seasons’ products. By establishing these relationships between the new and old products, machine learning determines the new products that should replace previous seasons products. The attributes have priorities such that the lineage must share some attributes, e.g. gender and product type, and key attributes, e.g. material and activity, have more influence than non-key attributes, e.g. designer and facility.
Similar products are the products that aren’t direct replacements (i.e. product lineage), but still share numerous product attributes, along with the same priority as for the lineage. If a lineage product does not align with the needs of a particular customer, these similar products can be bought instead. For example, a cotton blue dress with a round neck that is $25 is similar – and could potentially be an alternative – for a cotton blue dress with a v-neck that’s $30.
Complementary products are add-on products that are bought by most other retailers with a particular product. When customers are ordering products, these complementary products are often added-on to the order. For example, most shoe retailers typically order soccer balls with soccer cleats, where the soccer balls are an add-on sale.
The automated personalized offer is the machine’s prediction of what each retail customer will purchase based on the assortment plan, product lineage, complementary products, and the products purchased in previous seasons. This offer is either fine-tuned by the sales rep or presented directly to the retail customer.
To break this down further, the purchased products from a previous season are replaced with the current season’s products based on the product lineage, provided that the current season’s products are in the assortment plan. In addition, if a complementary product is very strongly related to their offer and part of their assortment plan, it is automatically added to the offer.
Machine learning Applied to the Seasonal B2B Sales Process
These product relationships are surfaced to the sales agent, along with highly relevant and comprehensive customer insights – enabling the salesperson to sell more efficiently throughout their entire sales process.
At 4-Tell, we are working with clients to optimize the usage of our machine-learning insights and envision that our technology will be applied throughout the sales process as follows:
- Pre-season sales preparation – machine learning creates the automated personalized offer for each retailer.
- Pre-season sales – machine learning performs gap analysis by identifying gaps between the retailer’s orders and the personalized offer, notifies the seller and customer of a gap, and suggests similar products to fill identified gaps.
- Entire sales process – machine learning sends notifications of non-manufactured products and suggests similar products to replace these dropped products.
In all three steps, the sales rep can monitor and fine-tune the automated suggestions, e.g. for mid-size retailers, or accept “as-is”, e.g. for small retailers.
Optimizing the sales process for mid and low-tier customers in this way allows for sales reps to put more time toward the customers that drive significant revenue, while providing more impactful personalized offers for lower priority customer that take less time to surface.
4-Tell’s Smart Commerce℠ Platform
4-Tell’s Smart Commerce℠ Platform has these technical capabilities to optimize sales efforts and drive additional revenue for businesses – and we make it easy to integrate.
4-Tell’s platform integrates with businesses backend data, B2B website, and assortment and inventory tools, such as the ERP. The platform does not require that the website and ERP to be integrated. The platform pulls catalog, sales and customer data feeds into a SaaS solution, and uses a REST API for integration into the website and ERP. This ensures that our integration is complete in a matter of weeks. Not years.
Considering the value of our product and our easy integration, businesses have no reason not to implement this solution into their sales teams. 4-Tell’s sales enablement platform is ready to deploy in your workforce. What are you waiting for?