Voice or speech analytics is still surprisingly new for housing providers, though widely used in the private sector. The latest Babel (2016) contact centre report, The UK Contact Centre Decision-Maker’s Guide, states that only six per cent of the housing sector currently uses voice analytics. But with 19 per cent actively planning to do so, the sector’s interest in customer analytics is growing.
All call centres hold customer-rich data. Almost all the business information we could possibly want about our customers is already coming through our call centres, so why would we not use it to sharpen up our business and improve the customer interface?
At Soha Housing, we have been using this clever software since 2015 and, with a combination of training, trial-and-error and intellectual curiosity, we have established ways to help us source issues, search voice recordings, embed interventions and instigate faster recoveries. By investing in technology, we use it to reduce repeat calls, resolve complaints earlier and build ‘right first-time’ responses.
With customer satisfaction levels at 92 per cent and costs low, Soha’s reputation as a landlord is solid. Our staff have a reputation for getting things done and understanding our services.
Our ’voice of the customer’ programme doesn’t mean that customers get what they want at any cost, but that the business gets what it wants and needs, because it gets closer to the truth about a service faster to prevent business lag. How? By searching an index of 30,000 calls within minutes, rather than waiting for survey results. Calls are uploaded overnight so all information is current.
Yes, the customers are often pleased and impressed with faster action, but the principle driver is business efficiency in real time, alongside customer satisfaction; we don’t see this as an either/or situation.
Most customer interaction analytics solutions use speech engines that are either phonetic or speech-to-text. In our system, all calls are phonetically indexed to allow the analysis to take place and we have no dependency on a language model and dictionary to identify words correctly. This allows us to access the entire content of calls, not just initially specified keywords and phrases as needed to enable root-cause analysis. Identifying clusters of terms gives us a starting point for deeper analysis.
We can search the rolling list of calls downloaded overnight and act on the results. The oldest calls are automatically wiped as new calls are loaded. Business interruption is minimised and interventions are instigated upstream in the business process.
With our customer service advisors handling anything between 250 and 500 calls per day, we know they are vital to the business. While they can recall and recount some of the best and worst situations, we really want to hear the issue from a range of callers to elucidate and validate the issue into either a one-off, a pattern or, at worst, a trend; we are looking for ‘runners, repeaters and rarities’ so that call volumes can be reduced.
Finding that one of our grounds maintenance team was struggling 10 days into the grass-cutting season last year saved us weeks of complaints, and we were able to liaise with the contractor concerned to get the service quickly back to a premium level within days, significantly reducing customer attrition.
There is certainly a place for traditional surveys but in comparison to searching recorded calls or waiting for results to be analysed, analytics gets us there much faster and with the irrefutable evidence of customer feedback. Importantly, any dissemination of findings is redacted and anonymous.
So who does the searching? We put together a team of operational managers who know what they are looking for so that the impact on the IT team is minimised.
There is undoubtedly a discipline needed; listening to the calls at the point highlighted where the keyword appears, and honing searches by customer phrases rather than tech-speak certainly unlocks more evidence. Maintaining the momentum also drives success.
Nexidia, the company who provide the voice analytics software, trained our team to establish call searches that reflect the customer conversation so that “the contractor didn’t turn up” takes us straight to the times, dates and locations at the relevant stage in the call.
When we find a problem, the intervention or ‘fix’ is identified with input from the operational managers to get to the root cause and, importantly, a sustainable solution. We are gradually driving out the seasonal service issues, so they don’t reappear as a new problem to a new manager a year later. Nexidia maintains that interaction analytics isn’t something you buy, it’s something you do. We completely agree!
Since implementing the Nexidia software, we have centrally recorded all issues and resolutions we’ve identified to demonstrate the return on our investment. These service improvements are often linked to improving customer care or efficiency savings.
Understanding customer sentiment is crucial to understanding customer behaviours. Hearing how welfare reform affects tenants as it’s rolled out and how our teams are responding has helped maintain our continued low arrears of 1.9 per cent. For example, understanding why some callers were reluctant to pay rent online helped to fine tune our approach to channel shift.
Finally, a level of trust is needed. With call-handler conversations now under scrutiny for breaks in a seamless service, we needed to make sure we avoided blame and instead recognised the good practice alongside the learning points for individual staff. Calls are quality assured to better understand talk periods, silences and call durations. We also check our customer services team are adhering to scripts to help ensure a consistent and efficient approach to customer care. We have a coaching matrix and know which caller can be called on for which skill – training, negotiating, problem solving, technical skills or empathy.
Our involved customers love the directness of the process. They know that we can hear the emotion of a call where someone with a young child needs a new home or a tenant of 30 years finds themselves with a difficult neighbour. Any significant issues that remain unresolved are taken to a meeting with the CEO who looks at the structural issues to decide if a change is needed in the business.
The customer experience tells us about our changing world and we see our business as a joint enterprise with valued customers. Their voice matters. That way we are all satisfied.
Maureen Adams is director of customer services and operations, and Lee Hayward is assistant director of customer services at Soha Housing.
