Turning a Cost-Focused Contact Centre into a Revenue Engine. An AI agent assist contact centre case study.
A UK soft furnishings distributor worked with Trimontium ai to unlock commercial value from everyday customer conversations. By deploying real-time AI agent assist and conversation intelligence, they turned missed moments into measurable results, boosting upsell conversion by over 25%, cutting operational cost through fewer repeat contacts, and improving customer satisfaction scores for the first time in two years.
Impact: Increased upsell conversion by over 25%, reduced operational costs through fewer repeat contacts, and improved customer satisfaction scores for the first time in two years
Case study: Deploying real-time AI agent assist and conversation intelligence to unlock commercial value from existing customer interactions
Client sector: Soft furnishings distributor
The brief: To transform a cost-focused contact centre into a revenue-generating asset by identifying and acting on commercial opportunities within everyday customer conversations, whilst improving operational efficiency and customer satisfaction.
The Client
A well-established soft furnishings distributor supplying retailers, interior designers, and trade customers across the UK. With a broad product catalogue spanning fabric and blinds to curtains and accessories, their contact centre handled everything from order queries and stock availability to returns and bespoke trade enquiries. By any measure, it was a busy operation. By most internal measures, it was also an expensive one.
The Challenge
The contact centre team were good at their jobs. They resolved queries efficiently, kept customers informed, and maintained solid relationships with trade accounts. But the operation was entirely geared around resolution: closing queries quickly and moving on.
Leadership had a nagging sense that something was being left behind. With hundreds of trade and retail conversations happening every week, they were sitting on an untapped opportunity. Agents knew the products well, customers called in regularly, and yet there was almost no structure around identifying or acting on commercial moments within those calls.
Meanwhile, costs were creeping up. More SKUs, more complexity, more calls. The contact centre headcount was growing in line with volume rather than in line with value. And customer satisfaction, while decent, had plateaued. There was no systematic way to understand what was driving friction or delight in the customer experience.
Trimontium ai was brought in to help the business think differently about what its contact centre could actually do.
What We Did
We started with the data, not the technology.
Before recommending anything, we spent time analysing the existing call and chat interactions, looking at what customers were actually asking, how agents were responding, and where the natural commercial moments were being missed. The patterns were clear almost immediately. Trade customers calling about stock availability were rarely being asked about complementary products. Retail accounts placing repeat orders were not being flagged for loyalty conversations. And a meaningful proportion of contacts were repeat callers chasing the same unresolved issues, a friction point that was quietly eroding satisfaction.
Then we deployed real-time agent assist.
Working with the client’s existing contact centre platform, we integrated Trimontium ai’s real-time assist layer, via TriMetrics giving agents live prompts during calls based on customer history, purchase behaviour, and the content of the conversation itself. When a trade customer called about a fabric order, agents now received a contextual nudge: “This customer hasn’t ordered from the new Autumn collection. Consider mentioning it.” When a retail account called for the third time in a month, a flag surfaced immediately: “Repeat contact. Check for unresolved issue before proceeding.”
Critically, we worked with the client’s team to make sure these prompts felt like helpful information, not a sales script. The tone and framing mattered enormously in a relationship-driven industry like soft furnishings.
We built a conversation intelligence dashboard
Every week, the client’s contact centre manager now has visibility across key themes emerging from customer interactions: which products are generating the most confusion, which stock issues are driving repeat contacts, and where agent responses are diverging from best practice. This was not just useful for the contact centre. Product and merchandising teams started attending the monthly review, because the data was surfacing genuine commercial intelligence they could not get anywhere else.
We redesigned how performance was measured
Together with the client’s leadership team, we replaced a dashboard that was almost entirely cost-focused with one that balanced efficiency with commercial contribution. Agents were no longer measured only on handle time and resolution rate. Revenue influenced, upsell take-up, and post-call satisfaction scores all became part of how individual and team performance was understood and rewarded.
The Results from our AI agent assist contact centre case study
The changes did not happen overnight, but they were measurable within the first quarter of full deployment.
Revenue and upsell improved materially. With agents better informed and supported in the moment, upsell conversion on inbound trade calls increased by over 25%. Agents were not selling harder; they were selling smarter, with the right prompt at the right time making all the difference.
Operational costs came down even as call complexity increased. Because repeat contacts fell significantly, driven by better first-call resolution and earlier identification of recurring friction points, the pressure on headcount eased. The team handled a higher volume of useful conversations without a proportional increase in staffing costs.
Customer satisfaction scores moved in the right direction for the first time in two years. The improvement was most pronounced amongst trade customers, who reported feeling better understood and better served. Post-call surveys highlighted agent knowledge and responsiveness as the standout positives, a direct reflection of the real-time assist doing its job quietly in the background.
What the Client Said
“We always knew there was more value in our contact centre than we were extracting from it. We just did not have the tools or the framework to get at it. Trimontium ai helped us see the operation differently, and the commercial results followed. Our agents feel more confident, our customers feel better served, and our leadership team finally has the data to make good decisions.” – Contact Centre Director.
The Takeaway from our AI agent assist contact centre case study
Soft furnishings is a relationship business. Customers, whether trade buyers or retail accounts, expect to be known, understood, and well-advised. The contact centre, when it is working well, is exactly where that relationship lives.
What Trimontium ai helped this client recognise is that serving customers brilliantly and driving commercial value are not in tension. They are the same thing. The technology just makes it possible to do both, consistently, at scale.
Interested in what this could look like for your contact centre? Get in touch with the Trimontium ai team.
