When facing the prospect losing the technical and institutional expertise of its founder and MD Denver Coleman, Polyflor South Africa went looking for a technology that could help. They came upon CLEVVA, a virtual advisor platform that enabled it to capture Coleman's 55 years of experience on an AI-powered platform.
Polyflor SA is a Johannesburg-based vinyl flooring and building protection solutions company which services architects and commercial clients within the healthcare, commercial, education, retail and hospitality sector.
The AI-powered platform was developed by Stellenbosch-based technology company CLEVVA, using its decision navigation technology. The CLEVVA platform allowed Coleman and his technical management team to capture expert logic into the system and then make it available to staff as they perform everyday duties, via an app named PolyPro.
Coleman says it only made logical business sense to deploy the intelligence augmentation technology within the company.
"I've been in the business for 55 years and a few years ago I attended an industry investment talk where a global specialist highlighted that if a business is not involved in some computer management or marketing in the near future, one would most likely not be in business anymore.
"For us this was a way of risk-proofing the company from losing valuable knowledge as well as enabling us to remain relevant in this rapidly digitising world. We now virtually have our own internal 'Google-like' personal database, which entails extremely detailed technical sales manuals which are accessible to the team at any given time."
Tandy Coleman, incoming CEO as well as Denver Coleman's daughter, says the deployment of the system has been a work in progress. "The journey began five years ago and since then we have been able to refine, add and remove certain parts to have it function to the best of its ability. The system is constantly changing as we add more information to it."
Developing the digital expert
Ryan Falkenberg, co-CEO of CLEVVA, says the development started with Polyflor's existing systems.
"The starting point was the CRM functionality that managed the efforts of all of Polyflor's sales and service staff. This system ensured that staff activities and customer information would be managed effectively. We then defined all the key decisions he [Coleman] knowingly would be asked to make, and unpacked all the factors he would consider in making these. Because the system takes a data - not decision-tree - approach to logic capture, the focus was more on the decision outcomes and the factors that had to be considered in reaching these outcomes. In effect, all logic was captured first into Excel sheets and then uploaded into CLEVVA.
"Then came the part of augmenting this system capability with a digital expert that could guide staff through the 'how' of every required activity. By adding digital intelligence to standard CRM functionality, we removed the need for staff to be highly trained in all the products, processes and technical aspects of the business. PolyPro effectively offered everyone a digital expert at their fingertips that would help ensure they consistently asked the right questions, offered the right answers and took the right actions, without them having to know it all. Additionally, they were then automatically given the detailed report back into the CRM system to close the loop."
Coleman adds that in the event that the system is unable to give an accurate answer, it will route the question through to one of the technical experts and he/she will then add the answer and save the recommendation into the digital expert system, allowing them to increase the accuracy daily.
"The trick is more in adding to the variables and factors that need to be considered and may have been forgotten in the knowledge capture sessions. So it's an iterative process. We capture, build and then test. We then pick up variables we did not consider and then add these quickly and retest. Over a short time frame, the logic considers every known factor and reflects the collective thinking of Denver and other sales and service experts within the team," explains Falkenberg.
He says that the human adoption aspect, on the other hand, proved to be quite a challenge. "The biggest challenge was getting staff used to the idea that their value was more in offering an amazing customer experience than in their specific product or technical knowledge. Another challenge was building the internal capability of capturing and maintaining complex expert logic - unpacking the expert logic in the first place, plus fine-tuning the upfront data. It starts with targeting the key areas first, and then adding to the expert database on a daily and weekly basis. Over time the digital expertise builds in scope and power."
Despite the challenges, Tandy Coleman says the system has allowed them to scale. "Previously we'd hire people from within the flooring sector, as this would speed up the on boarding process. However, by relying on product knowledge, these salespeople generally reverted to learned-habit selling, and not identifying their clients' needs.
"It was a bit of a catch-22, because to offer advice you need the product and technical knowledge but to gain this knowledge takes years. With the PolyPro, we can provide our staff with instant product and technical expertise. This allows them to focus their energy on really understanding our clients and their needs, with the confidence that their digital expert will help shape the solution based on all the different factors and requirements. Our team can specialise in the customer experience, while the digital expert supports them with product expertise."
She adds that the company is looking to integrate the PolyPro into its Web site to enable a self-service portal for clients.
Share