In a recent IT leadership meeting, someone asked how we get our CFO and other leaders to prioritize work that improves consumer experience (CX) but does not have an associated hard ROI.
Business leaders are obviously genuinely interested in driving those CX improvement efforts and are well aware of the challenges in prioritizing them. The onus is on us to help them make apples-to-apples comparisons when rationalizing our strategic project and product feature portfolios, which will help us get those CX projects above the line.
Here’s a carrot: a November 2023 Gartner report concluded that businesses who have calculated the relationship between customer satisfaction and business outcomes are 29% more likely to report CX-related budget gains.[1]
But as we all know, consumer-centricity is relatively new to the business of health care. Therefore so is our measurement of the value of improving our customers’ perception of us. Which means we have work to do.
The measurement of customer experience per se has become mission critical for payers. For example, in the 2020 MedAdvantage STARS ratings, member experience measures were only 31% of the weight. In 2023 that grew to 58%.
What we need now is to establish a demonstrable link between customer experience metrics and revenue.
In 2019 Gartner conducted a study of 249 business decisions[2] that affected customers and were based on complex circumstances. They concluded that the key to the CX / ROI dynamic is understanding how what matters to specific customer groups pushes specific operational levers of profit and loss – and that operational leaders, with their deep understanding of the consumer, are a key resource in making the right calculations.
The standard model Gartner identified was to first determine the outcomes impacted by CX, then gather CX metrics such as Net Promoter Score (NPS) and Customer Satisfaction (CSAT), and finally do regression analysis to correlate changes in metrics to changes in outcomes, and in turn revenue.
Clearly one of the biggest impacts for a payer is churn at open enrollment. The overall equation there is easy: add members == good. At a finer grain, decrease overall utilization per member == good.
Here is a naïve illustration using NPS[3] – the numbers are all made up by way of example. First we take our made-up NPS score of 54, break it down into its constituents, and map to our membership.
| NPS Category | Percentage of Members |
| Promoters | 68 |
| Passives | 18 |
| Detractors | 14 |
Now let’s look at open enrollment churn, again with made up numbers that assume promoters are more likely to re-up than detractors.
| NPS Category | Percentage Retained at Open Enrollment |
| Promoters | 90 |
| Passives | 70 |
| Detractors | 50 |
If we do the math, the percentage of members retained at open enrollment would be 80.8%.[4] If we assume we have 2M members, that would be 1.616M retained.
Now let’s increase our NPS by 10 points. The NPS is promoters minus detractors. So leaving Passives the same, we move 5 percent from Detractors to Promoters, giving us these NPS category proportions:
| NPS Category | Percentage of Members |
| Promoters | 73 |
| Passives | 18 |
| Detractors | 9 |
If we do the math again, with the new values, we would retain 82.8%, or 1.656M, a gain of 40K members. If each member represents $12K in revenue, that would be a swing of $480M dollars. Each NPS point gains us $48M – in this hypothetical example only.
That is a naïve example for several reasons, most notably that in healthcare we have complex customer relationships. Most payers have relatively few individual members. The overwhelming majority are from groups. So the satisfaction of individual members in those groups has only an indirect bearing on churn – their dissatisfaction has to bubble up to group leadership and persuade them to revisit their contract.
Group satisfaction, as opposed to member satisfaction, would seem to be the key metric we would want to correlate with churn.
At a high level, Forrester has a metric called the Customer Experience Index. They publish a yearly report[5] which includes calculations by industry for the incremental increase in annual revenue per customer of a one percent increase in the index. Those range from $0.08 for credit card providers up to $104.16 for luxury auto brands. They don’t include health care or insurance, but auto and home insurance comes in at $14.32 per point.
Others have done calculations based on NPS by industry, and overall, such as a London School of Economics study from 2005 which found in general a 7 point increase in NPS correlated to a 1% increase in revenue[6].
Gartner cautions against applying a standard calculation without regard to context:
Standard calculations assume that each CX and loyalty metric holds the same weight as others and that the benefit of improving enterprise CX is the same no matter how good the experience already is. Unfortunately, the value of improved CX may differ greatly in moving from poor to acceptable versus moving from good to great. While the standard model is compelling in its simplicity, the lack of nuance actually diminishes the likelihood stakeholders will accept its output at ace value.
So how do we make progress?
Some of us are already measuring NPS. MedAdvantage payers are also producing AHRQ Consumer Assessment of Healthcare Providers and Systems (CAHPS) scores, which roll up into STARs ratings.
We can connect the dots pretty readily in Med Advantage, where the recipe for the impact of improved CAHPS scores on STARs is well-known, and we can estimate the value of gaining – or losing – STARs.
There are also revenue calculations for some narrower, less direct satisfaction metrics such as First Call Resolution (FCR) in Customer Service call centers, where you can calculate the reduction in customer service agent time, and by extension cost, associated with improvements in FCR – which we assume improves customer satisfaction.
In general, rigor of the forecast ROI comes from establishing a statistical correlation between a CX metric and an operational measure, probably using some sort of multivariate regression analysis. That means you need historical data on both metrics and measures so you can ‘mine’ for those correlations. If you don’t have, or can’t retroactively determine, those operational-lever correlating metrics with usable frequencies, you may have to invest now in more and better metrics to enable correlations down the road.
Here is the promised suggested path:
- Partner with Strategic Finance, for their expertise in modeling and forecasting, Operations, for their expertise in operational levers, and your Quality Management folks for their expertise in metrics.
- Work together to identify key operational levers which affect revenue.
- Identify the consumer cohorts associated with those levers.
- Figure out the appropriate metrics for those consumers’ satisfaction in context.
- Make sure you have time series data for both lever and meter.
- Analyze the correlation between the CX metrics and changes in associated outcomes.
- Calculate the revenue impact of CX improvements per those metrics.
- Forecast project or product feature impacts on those metrics and generate the expected ROI.
For your initial scope, drill down in collaboration with partners on a specific example or two to work through the challenges.
You might also consider bringing in a consulting firm with expertise in the area to help guide you. Bain would be a good choice if you can afford them – they are the O.G. of NPS consultancies (hey Bain, how about a free t-shirt – a roomy 3XL in 100% cotton if you can manage it : ).
Stay tuned.
[1] Gartner, ‘Prove the ROI Business Case of Customer Experience Programs While Staying Customer- Centric’, Augie Ray, 10 November 2023
[2] Gartner, ‘Calculating the ROI of Customer Experience Initiatives’, 2019
[3] Adapted from Chapter 5, ‘How to calculate the impact of ROI on revenue (ROI of NPS), from Thematic Insights ‘Calculating the ROI of CX’
[4] (0.90 * 0.68) + (0.70 * 0.18) + (0.50 * 0.14) = .808
[5] Recent Forrester reports are proprietary, the data here is from a 2016 report posted by Pointillist, a customer journey platform vendor.
[6] ‘Advocacy Drives Growth’, Brand Strategy, London School of Economics, Dec 2005.
Health IT Architect 