Small and medium-sized health organisations are often drowning in data but lack actionable insights. The real issue is not access; it's that many organisations are not linking data analysis to commercial questions, or creating a feedback system that turns scattered data into valuable insights over time.
When I come into an engagement, the data audit is one of the first things I do. Not the CRM audit, not the pipeline review, the full data audit: every system that captures information about clients, stakeholders, service use, feedback and commercial activity. What exists, what is missing, what is being collected but not used, and what the gaps in the data are telling us that the data itself cannot.
What follows is a plain account of what that audit typically surfaces and why this can present significant opportunities for growth and sustainability.
What the data usually shows
Health organisations at the SME scale tend to have more usable data than they think and less structured data than they need. The most common findings across engagements fall into five categories.
1. Service and product utilisation patterns. Records of what clients are actually using, how frequently and at what intensity exist in most systems, but are rarely analysed for churn signals. A client who has reduced engagement over 90 days is a churn risk. A client who has never used a core feature is a support failure waiting to become a cancellation. Neither shows up as a problem until the client is gone, and by then, the data that would have predicted it has been sitting unread for months.
2. Customer fit. Most health organisations have grown their client base opportunistically. They took the clients that came to them rather than systematically building toward a defined profile. The data almost always reveals that a small proportion of clients generate a disproportionate share of revenue, require less support, renew more reliably, and refer others. Those clients share characteristics. The organisation usually has enough data to identify those characteristics and build an acquisition and retention strategy around them. Many of them have not done it.
3. Archetypes built from actual behaviour. The academic customer persona exercise, in which a team sits in a room and describes a fictional ideal client, produces archetypes that reflect the team's beliefs rather than the data. Behaviour-based archetypes built from real utilisation data, support patterns, communication preferences and contract histories are different. They are more specific, more actionable and more likely to predict what will actually work in outreach and product development.
4. Customer-facing system failures. Slow response times, broken workflows, billing or service delivery errors, and gaps in communication all generate signals in the data. Those signals are often not monitored effectively and systematically at the SME level. The result is that preventable failures become repeating failures because nobody connected the incident log to the commercial consequences.
5. Product development feedback loops. Whether the product or service development team is actually implementing what clients are telling them is visible in the data if you know where to look. The gap between what clients request and what gets built, the time between feedback and response, and whether resolved issues reduce or repeat are all measurable. In many organisations, they are not being measured.
The complaints problem nobody wants to talk about
Of all the data gaps in health organisations, the feedback and complaints infrastructure is the most consequential and the least addressed.
The reluctance is understandable. Complaints feel like failure. Most founders and operators would rather build a better service than acknowledge that dissatisfied clients exist. What they do not understand is that the lack of a structured complaints system does not mean there are no dissatisfied clients. It means dissatisfied clients are not telling you.
Research by TARP, the Technical Assistance Research Programs group commissioned by the White House Office of Consumer Affairs, found that for every client who makes a formal complaint, approximately 26 others with the same problem say nothing. They do not complain. They leave, or they stay and disengage. And each of those silent dissatisfied clients tells between 9 and 15 others about their experience (TARP, 1986, 1999).
The mathematics of that is not comfortable. One visible complaint represents a substantially larger problem that is already circulating in the market. In health, where referral networks are tight and professional reputations travel fast, the reputational cost of unmanaged dissatisfaction is higher than in most sectors.
What an ISO 10002-conformant complaints system does is create a structured channel that makes it easier and safer for clients to raise concerns directly, before those concerns become market-circulating negative word of mouth. Research by Ang and Buttle (2012), examining 144 organisations across four industry sectors, found that ISO 10002-compliant complaints-handling processes account for 24% of the variance in customer advocacy and satisfaction outcomes. The mechanism is straightforward: when clients know there is a genuine process for raising concerns, more of them use it, fewer of them leave silently, and the organisation gets usable data about what is actually going wrong.
Most SME health organisations do not have this infrastructure. They have a contact form, possibly a satisfaction survey, and an inbox that someone checks when they have time. That is not a complaints system. It is the appearance of one.
The distinction is critical because a genuine complaints system is not primarily about handling complaints. It is about capturing the signal that tells you where your service, product or relationship is failing before the client decides not to renew, not to refer, or to tell their professional network why they moved on.
What is missing from the data
The audit almost always surfaces gaps that are as informative as the data itself.
No mechanism for capturing client satisfaction at defined intervals, not at contract signing and not at annual review, but at meaningful points in the service relationship. No correlation between operational data and commercial outcomes: support tickets are not linked to renewal rates, billing errors are not linked to churn, and feedback responses are not linked to referral patterns. No baseline data against which to measure whether things are improving or deteriorating. And no structured way for product or service development teams to demonstrate that client feedback has influenced what gets built or changed.
These are not difficult problems to fix. They require decisions, system configuration and the discipline to maintain the process once it is in place. What they require most is someone who knows which questions the data should answer and can build the infrastructure around those questions rather than around what is easiest to measure.
What changes when you actually use it
Organisations that build proper data and feedback infrastructure do not just reduce churn. They change how they acquire clients, develop products, engage existing relationships and allocate commercial effort.
When you know which clients generate the most sustainable value and what they have in common, acquisition becomes less speculative. When you know where service failures are recurring, product development has a priority list driven by client feedback rather than internal assumptions. When you know who is at risk of churning before they churn, retention effort can be directed at the right relationships at the right time.
None of this requires sophisticated technology. It requires the right questions, the right data infrastructure, and someone who understands both the commercial logic and the sector well enough to know what to look for.