Using RPA to Process Quotes and Improve Customer Experience
If you’re working in B2B sales, and you’re managing volumes of thousands of RFQs (RFQ = Request for Quotation) per month, automating the quoting process and being able to respond to your customer in under one hour can significantly improve customer experience during the sales process. In this article, we will explore some of the solutions that RPA enables, and map them to the steps required in a quoting process:
Gather data: Extract product features from the tender
Identify products: Which products are being requested, and in what quantities?
Read customer profile: Are there specific deals in place with this customer that will affect prices?
Generate offer
Review & send offer
Sample RFQ processing flow
Let’s dive deeper into these process phases and take it step by step.
1. Gather data
Let’s say you are receiving RFQs via e-mail. There are two information categories that you need to extract:
Customer Information (name, address, contact, etc.)
List of requested products (line by line)
UiPath’s machine learning powered Document Understanding Framework has a generic ML model that can be trained to extract these pieces of information, from the specific kind of RFQs you are getting on a regular basis. The power of ML enables this model to be flexible enough to handle varying inputs from your various customers, as long as they are not radically different from one another and follow the same patterns (e.g., order items organized in a table like format).
2. Identify products
Now, that you have all the order lines extracted, you need to match them to your internal catalog from SAP/other ERP system, and make sure that the bot knows which exact product is being referenced. This is fairly easy if the RFQ contains a product code – the robot will be able to search for the code format (e.g., 12 digits). If, however the table only contains a product description (e.g., Exterior PVC double door - 136x226), the robot will have to learn the keywords that correspond to each product, as well as extract dimensions and other features. Then it will use all these parameters to search in your product catalog and identify the matching products.
Data extraction & Excel product catalog matching sample
3. Read the customer profile from your CRM
Having all the requested products identified, the next step is to prepare the quote. As you might have some discounts negotiated with the customer, you will need to read these from your customer’s CRM profile. You already have the customer's name extracted in step 1.1, so you can easily identify the right profile. Then you just apply the discounts based on product, product category, or overall discount and you have the final calculation for the quote.
4. Generate offer
The easiest way to generate the final document is to use a word template with your branding, terms and conditions and other predefined content. This template will have placeholders for the customer’s name, dates, and a table that the robot can extend with as many rows as needed.
Alternatively, you can do this directly from your CRM, as highlighted in the example above.
5. Review & send offer
In the final step the account manager for the given account is notified that there is a new quote pending. The AM reviews the prepared quote, makes any final adjustments and sends it to the customer via e-mail.
This process is just an example of a bot that we built to support processing large volumes of data from various systems while humans maintain control of the end result. For more examples of how RPA can further improve your sales operations, you can continue reading our article “How RPA Supports your Sales Team” or watch our recent sales webinar on-demand.
