How do you turn customer feedback into an evidence-based action plan?

What is the Wordnerds Customer Feedback Framework?

The Wordnerds Customer Feedback Framework is a three-stage method — collect, categorise, prioritise — for turning unstructured customer feedback into evidence-based action plans. It begins with the core objectives and key metrics a business already tracks, such as TSM scores, customer satisfaction and complaints, then works through the data channel by channel. Wordnerds built the framework by observing what the best insight teams across hundreds of organisations do consistently, regardless of sector. Its design goal is to serve teams who feel "data-rich but insight-poor": overwhelmed by volume, without a clear way to prioritise, and lacking the evidence to support an action plan. Helen Precious stresses that you do not need a fully integrated feedback ecosystem to begin — no organisation has one — so siloed data in spreadsheets is a legitimate starting point. The framework continues beyond prioritisation into an impact loop (act, measure, prove), which Wordnerds covers in follow-up sessions.

How should you categorise customer feedback at scale?

Categorisation is the stage where Wordnerds says most of a feedback programme's value is created — and where most teams get stuck. The goal is to move from "chaos" (messy, unstructured comments) to organised data you can make decisions from. According to Helen Precious, a good end result has three properties: exact numbers (how many people raised a specific issue, not a vague summary), sentiment by topic (how people feel about each theme, not just overall), and the ability to drill down from broad topics to root causes and locations. The session is clear that manual tagging, while admirable, introduces human bias and inconsistency between colleagues and rarely keeps pace with rising volumes. The recommendation is to use technology to handle the volume while preserving the human element of what tenants are actually saying — because in social housing, the data represents people's lives, not just scores.

Why don't general-purpose AI tools work for regulated feedback?

General-purpose GenAI tools such as Copilot and ChatGPT are convincing but unreliable for regulated feedback analysis, according to Helen Precious. Three problems recur. First, they summarise without quantifying — they will say "some people mention damp and mould" but not that comments about repair quality rose by a specific percentage, and that specificity is exactly what the business needs. Second, they hallucinate: issues that sound plausible may not actually exist in your data, so every output needs checking. Third, many open large language models feed your data back into a wider training model, which is a serious problem for infosec teams handling sensitive tenant information. Out-of-the-box "black box" tools bolted onto survey platforms have their own flaw: you cannot see why they reached a result, and they are typically tuned for high-street retail, so they lack the sector knowledge to handle damp and mould, mental health or tenant vulnerability appropriately. Wordnerds advises staying away from anything you cannot interrogate.

How do you prioritise which feedback to act on?

Prioritisation turns categorised data into actionable recommendations and protects teams from acting on gut feel. Helen Precious explains that sampling data and making gut-feel decisions rarely moves the needle, because such decisions are not representative or evidence-based and get skewed by the loudest voices and recency bias — for example a CEO's pet question becoming priority number one. The framework instead works through a ladder of techniques: volume-and-sentiment analysis to surface what people talk about and how negatively; root cause analysis to follow an issue down through its layers; driver analysis to explain why a satisfaction score moved; and, for the statistically confident, Pareto analysis and random forest models. Pareto is highlighted as especially useful: it tells you that fixing three specific things will drop complaints or lift sentiment by a measurable amount, turning a list of issues into a ranked plan. The best teams, Helen notes, use multiple frameworks depending on the objective rather than relying on one.

What frameworks does Wordnerds use for social housing feedback?

Wordnerds breaks feedback down through "frameworks" — lenses that match the metrics and regulations an organisation already cares about. For social housing, Sarah Wilson confirms the most common is the TSM framework, a pre-trained theme bank of 126 themes built around the core social housing TSM categories, which aligns every survey to numbers teams can trust. Analytical frameworks teams can apply today include user-journey analysis (mapping feedback to a repair or complaints journey), root cause analysis, Pareto analysis and COM-B, a motivational framework for understanding behaviour. Wordnerds also runs specialist work: a prediction beta with Town and Country that aims to distinguish correlation from causation and spot the triggers that escalate a grumble into a formal complaint early enough to intervene, and a new Awaab's Law compliance framework that maps feedback against the regulation to flag where an organisation may not be compliant. For employee-feedback work in other sectors, Helen Precious references the Hertzberg methodology.

How do you prove your feedback actions are working?

Proving impact is the second half of the loop — and, Helen Precious argues, where teams should spend most of their time. The framework deliberately stops at "prioritise" because once you have the right data and the right priorities, a new set of problems begins: convincing teams to act, and proving the actions moved tenant satisfaction. The aim is to shrink the time spent on the insight loop so teams can spend more on CX delivery, project management and measuring impact. The session points to results: at Granger, Jenny moved categorisation from days and weeks to minutes and now spends 60% of her time on the delivery side, and at Town and Country, Theo used the TSM framework to quantify complaints about workmen using dirty water to clean windows, prompting the business to install closer taps and clean water — with a measurable lift in satisfaction. Wordnerds also launched a five-minute maturity assessment across six categories, returning an overall score and a personalised next-step report.

Full Webinar Transcript

Sarah Wilson: Good morning, and thanks so much for joining us. We're always amazed that we actually get people to these things — you're not supposed to say that, of course.

Is it your first webinar? Certainly feels like it. Morning, Millie. How are we all doing? Have we all got a heat wave? Gateshead is feeling very sweaty right now, so let me know what it's like in your area and how you're coping in the school holidays. You shouldn't complain about the heat, really — it doesn't come around often enough for it to be a problem, but we don't have aircon, that's the main problem. I do say that whilst drinking a hot coffee.

I'll give everyone a minute to join. Let me know if you've got any nice plans for the rest of the school holidays. I made the mistake of taking a holiday in June, so now it's just a long slog to Christmas.

Right, I think people have joined. I'll give everyone one more minute and then we'll get kicked off into the good stuff.

So, thanks so much for joining — it's an absolute pleasure to have you on the call. Hopefully you're in the right place. We're going to be looking at the customer feedback framework, and then showing you that it actually works with some theoretical-to-practice examples of it really working.

Today you've got me — a very familiar face, I'm sure, probably sick of hearing from me. I work in the account management team, specifically in social housing. And then I've brought in the big guns today: my boss, Helen, who's freshly back from maternity leave. We thought we'd welcome her back with another webinar, just what she needs. So thanks for joining us, Helen, we appreciate it.

For today we've got roughly an hour planned, and then you can go away and get your lunch. We're going to start with some setting the scene — that's me — and then move into Helen, who's going to cover the framework and give you some examples of that in action. Back to me for a little bit of chat about this actually working with a specific case study, and then we'll go into next steps and Q&A.

Do put your questions in the group — we want to try and have that community feel and have your burning questions answered. Especially if they're difficult, we can hand them over to Helen. We've got a brand new resource we haven't shared before, so I'm looking forward to getting that out and getting some feedback on it.

We thought we'd kick off with a little poll. This is all about your feedback, and we've got quite a few new people on the call, so we were really keen to understand what your most pressing feedback challenge is. You can see everybody's answers as well, which is quite nice.

We've got a bit of a split so far. Too much data — yeah, I think we've been talking about data-rich, insight-poor for a long time. And "don't have a way to prioritise actions" and "no evidence to support action plan". Nobody feels like they're struggling to get buy-in, which is a nice way to start. "No evidence" is good, because Helen's going to touch on that. That leads us in nicely.

So, who are we? Just for the newbies on the call: we are an enterprise customer feedback analysis platform, specifically in social housing. What does that mean? It means we help you take your tenants' feedback — whether that's TSMs, complaints, transactional surveys, anywhere that you collect data — and get the action plan that you need to get buy-in to ultimately improve the customer experience.

I work specifically in the social housing space, so we've got some great brands — we work alongside Town and Country, Riverside, One Manchester, Guinness Partnership — a real wide range of both quite small and some of the biggest in the business. And then Helen deals with everything else: financial services, retail, travel and hospitality. What we've found over the years is that these brands all have very similar challenges and tend to approach the data in a similar way. So we're sharing best-practice learnings from them, as we always do at Wordnerds — we take their ideas and pass them off as our own.

Let me set the scene. One of the main things we're always told at Wordnerds is that feedback is just so challenging. There's a Chinese idea, weiji, which means that within every crisis or opportunity there's always a turning point. Feedback is super challenging — English in particular is very difficult, and dealing with words, who wants that? If only we could make them like numbers. But within the feedback there are so many opportunities.

We find the main problems are categorised into four points. First, your feedback volumes are always rising — tenants' voices are being amplified, there are campaigns to increase tenant feedback, you're getting more homes, there are mergers, and you're probably surveying more frequently. Second, regulatory pressure: Awaab's Law coming in October, the Make Things Right campaign — lots of pressure to get it right and to have that action plan.

Third, it gets really challenging when you're trying to wrangle this information across lots of complex systems. Anyone who's been through a merger knows how messy that gets. And at the same time, the number of people in your team is shrinking — some big housing associations are operating with two or three people in their customer experience team, asked to do more than ever with smaller budgets than we've ever seen. Staff need upskilling, but when do you have time to take that helicopter view?

Fourth, what does that mean day to day? You've got overwhelming amounts of data in lots of different silos. You can't categorise efficiently — what takes days should really take minutes. And you're stuck at surface-level insights when it's really hard to understand root causes and build evidence-based business cases that get buy-in. You want to change that "no clear robust feeling" to a proven framework that you can follow.

There's unlimited opportunity in verbatim. It gives us context behind the scores and metrics, shows us the real people, captures powerful emotional drivers and pain points, captures issues not covered in structured questions, and offers the authentic customer voice we're all trying to amplify. I hope that sets the scene. I'm going to hand over to Helen now to take us through this feedback framework.

Helen Precious: Thanks, and thanks for the reminder to come off mute. Hi everyone, really nice to be here this morning.

So this is the Wordnerds feedback framework. As Sarah mentioned, we sit with hundreds of teams every year, and what we realised quite quickly is that the best teams are doing very similar things in how they process and analyse their feedback, and how they turn that feedback into something useful, meaningful and impactful for the end customer — your tenants, depending on the sector.

This framework works for all sorts of challenges. It's designed to work for people who feel data-rich and insight-poor, which from the poll we can already see a lot of you are. The way it works is to take the core objectives the business cares about — typically in social housing, things like how do we improve tenant satisfaction, what's wrong with our repairs service, why has our TSM score changed, what evidence have we got to understand that change. Insight teams get inundated with all sorts of questions, but irrespective of the question, this framework is a nice way of working through those challenges and delivering something impactful — and giving you and your team the confidence that when you make recommendations, they're the right ones, they're evidence-backed, and the process isn't too onerous.

We start with your core objectives and key metrics — TSM scores, customer satisfaction. Then you collect the data, categorise it and prioritise it. For today, based on the poll, we'll focus on the insight loop — how do we get to a point where the insights are genuinely actionable, where your teams have a roadmap for fixing some of these issues. On a future webinar we'll give you advice on how to make those insights as impactful as possible once you've got them.

I thought I'd start with some common anxieties. Sarah and I sit on calls all year, and irrespective of the sector we hear the same things — this idea that you need to get your ducks in a row before you start. You don't need all of your surveys set up and running, touching every touch point. Most of you will have a TSM survey going out, some transactional surveys, and a complaints process. Everybody talks about the dream of all our systems being integrated, and I'm here to tell you that absolutely nobody has those systems set up.

At the same time, everybody thinks everyone else is further along than they are. Whenever we get on a call, most people are pleasantly surprised to realise they're not as far behind as they thought. And often what we see is "we're so busy, we'll look at this next quarter" — and the year's gone before you've blinked. So really the message is: if you've got verbatim feedback sitting in your remit — survey responses, complaints data, anything — you're ready to get started. You can make big differences even from a standing start.

So I'll talk you through the first three sections: how we collect the data, how we categorise it, and some ways to bring priorities and action plans into play.

Goal number one is to unify your feedback. The 360-degree customer feedback programme is probably the ultimate goal, the thing insight teams are told they need. What does that mean in social housing? How do we get all that customer feedback — anecdotal as well as formalised — into one place? TSM surveys, transactional surveys and complaints are the three we deal with most often. But what else could help inform what's influencing tenant satisfaction? Social media channels, CRM notes and call transcripts can house a lot of this, and your frontline teams who are face to face with tenants could feed in too.

There's complexity in unifying all this data, but what I'm here to tell you is that you can start from anywhere. We often talk about siloed data as a negative thing — it is a challenge, but it's often where most people start. If you're working in spreadsheets, collecting data through your surveys, that is the perfect place to start, because you're already starting the process. Look for your most accessible or highest-value data and start there.

At some point we want to start bringing in the next data set — it's a step-by-step process. Nobody says "here's all my data, go and analyse it all together." We work channel by channel. If we start with a TSM survey, how do those results compare to your transactional surveys? We look at the patterns, trends and comparisons, and then the next common one is complaints. The end goal is to think about how we integrate some of these systems — BI tools to report the data where the rest of your team are, data lakes to unify your conversations, bespoke integrations. That's a long process; by the time you're looking at an integrated ecosystem you're probably bringing in your IT teams, who can be stretched. So start with the siloed data. That's totally cool — we can do lots with it.

The most exciting bit is when we start categorising and prioritising. The aim of categorisation is to go from chaos — unstructured, messy data — into something more organised that provides clarity, so you can make decisions. This is the point where clients sit on one of two sides of the fence. Some love it — kudos to you; lots of analysts love being in the data, close to the customer verbatim, and it's really important we don't lose those human connections when we report. Particularly in your sector, these are people's lives.

The other people on the fence — shout out if this is you — are those manually categorising in spreadsheets, where people tend to want to throw their laptop out of the window. The process of manually going through all those comments is long-winded, and for a lot of people it's boring; it's a low-value task, and you're all high performers with more important things to do. Kudos to you if you're manually tagging, but particularly if you can't keep up with the volume, it's really important to look at ways technology can provide a lift to handle the volume.

Be aware that with manual tagging you introduce human bias, and coding with colleagues can be quite inconsistent — people's perceptions of what something is can be very different, so you might be missing things or not getting accurate numbers.

There's now a group of you trying GenAI — using Copilot or ChatGPT to summarise what's in the data. Whilst the summaries can sound very convincing, they often don't give you a number. They're good at saying "some people have a problem with damp and mould" but they don't say that comments on repair quality have gone up by 10% or 20 cases. That specificity is really important when you're analysing data and feeding it back to the business. The other thing — and if you haven't tested it, go and check — is that GenAI is really common for hallucinations, which is a fancy way of saying it makes stuff up. These issues may not actually be occurring in your data. And definitely check with your infosec teams: often these open-source LLMs use your data to feed back into a wider training model, at which point your infosec team will be very upset.

Then there's a group using out-of-the-box solutions — AI that comes off the back of your survey tools. They've started to introduce AI, but they often use what we call a black box solution: you can't necessarily understand why it gave you the result it did. We'd advise staying away from that, because you need control over the decisions it's making. Out-of-the-box solutions are also often designed for the traditional high-street retailer, so they lack specificity for the problems you see with tenant feedback. Customer service in retail is very different to social housing — they've never come across damp and mould or mental health in the way your data has, or the vulnerabilities your customers might have, so they don't treat those things the way they should.

When you're categorising data, the end result needs to put a number on things — you need to understand exactly how many people are talking about a specific issue or topic. We want exact numbers, whether on one spreadsheet or multiple channels. The second thing that's helpful is sentiment by topic: not only what people are talking about, but how they feel about it. That can be positive, neutral and negative, or something more detailed. And you want to be able to drill down to root causes — the top-level things people talk about (a repair issue, a safety issue, a customer-service issue), then the next level (within safety, a damp and mould issue, in this location), following the data down the rabbit hole to the point where you can actually action it.

The third thing is to prioritise this data. A few of you mentioned you have no evidence to support action plans, or no way to prioritise. The goal with prioritising is firstly to create actionable recommendations, and also to present the data in a way that makes sense to teams who get overwhelmed by new information.

Apologies if I'm still here — I'll turn my camera off, it's my wifi. Thank you, Linda.

So, actionable recommendations, presented in a way your teams understand. Often when you deliver insights they may not be wrong, they're just unfamiliar to your colleagues, so they don't know what to do with them. Everybody's busy, so we want to think about the things that make the biggest difference to your core objectives and key metrics.

We did a webinar on Tuesday, and when we ran the poll we asked: when you present customer insights, stakeholders usually — take notes and carry on as before; ask "but how do we action this?"; challenge the data or methodology; or share their own anecdotes that contradict you. All of these come down to how you present the data.

Where do people start if there's just too much data to handle? We often see people sampling the data and making gut-feel decisions. When you've been in the business a long time, those can be right on the money — but often they don't move the needle, because they don't represent your entire tenant group, they're not evidence-based, and you risk making the wrong decision. That's where insight teams feel worried about their recommendations, because if they don't make a difference to tenant satisfaction, it comes back to you. Gut feel also gets influenced by the loudest voices and recency bias — the CEO asks about something from a previous meeting and that becomes priority number one, even if it's not representative.

So the first thing we want to do is bring you to a point where you're at least looking at the data from a representative perspective — looking at volumes, and sentiment where we can, to make volume-based recommendations. A couple of ways to analyse that data: root cause analysis, going down through the layers to understand what's driving an issue; and driver analysis, looking at why a sentiment or satisfaction score went down last month. For the Excel wizards, pivot tables and cross tables looking at correlations between two sets of themes can be really helpful.

We've then got a group using statistical analysis — often from a data science team or BI analysts. Please don't sit there feeling you're doing a terrible job if you're not at this stage; it's a specific skill set and something we can help with. Things like Pareto analysis and random forest models look at correlations between your top-level metrics and the biggest influences, and a lot of this is done in BI tools rather than spreadsheets because it gives you more scope to model. If you're interested, we did a really nice webinar with Guinness Partnership where they broke down their BI dashboards, random forest models and decision-tree modelling for their real-life case studies. Let the team know and we'll follow up.

The way Wordnerds would recommend breaking this data down is by what we call frameworks. "Frameworks" might not be a phrase you use internally, but it's just a way of breaking your data down through the lens that matters most to your organisation. The starting point is: what regulations do you operate within, what metrics do people care about? Take the regulations — the obvious one is using the TSM criteria, mapping feedback from your TSM and transactional surveys through the lens of the TSM criteria: safety issues, maintenance and repair, where communication is fair and honest and transparent. Similarly, what metrics do people care about — CSAT, complaints — and a driver analysis for something specific like what's wrong with our repair service, breaking the data down by a repair journey rather than lumping everything together.

I thought I'd share some frameworks that work for our clients. The analytical frameworks are things you could work on today. User-journey analysis works nicely with your repair journey or complaints process — breaking down the steps to understand where issues arise and mapping your data to that journey. Root cause analysis — understanding drivers of complaints — most of you are probably already doing. Pareto analysis looks at the relationship between top-level scores and the data: essentially it tells you that if you fix these three things, the number of complaints will drop by X or your sentiment score will go up by X. It's a nice way of getting to prioritisation rather than just listing issues. The best insight teams aren't focused on one framework — they use multiple, depending on the location, objective or problem. COM-B is a motivational framework we use to understand how we influence different behaviours.

Then we've got specialist frameworks for our social housing clients. You don't have to use Wordnerds — we can share how we tag the data so you can break it down similarly. The first is a TSM framework: a series of categorisation models that organise your data strictly through the TSM model — the six core areas of the TSM survey. We feed all sorts of data into this, so it's used to understand not just the results of your actual survey but what data elsewhere — transactional surveys, complaints — could tell us what's likely to be an issue when we next run the TSM survey. As mentioned, the repair journey lets you understand data through a customer-journey lens, which is helpful for prioritising what to fix.

Then two new ones. The first is prediction — a beta we're running with Town and Country. Often teams look at statistical correlation, but correlation and causation aren't the same thing. We often see a correlation between the number of repairs and the number of complaints, but repairs aren't necessarily the cause of that increase. With the prediction model we're trying to better understand what causes a complaint, looking at the triggers and building AI models to understand what might cause an escalation to an official complaint — and if we understand those triggers early enough, can we intervene and reduce escalations. We haven't got the results yet; the theory is being tested, and I hope we'll have a webinar to share on it in the future.

And as Sarah mentioned, regulations are coming thick and fast in this sector. So we're also building frameworks that help you organise the data in a way that meets the regulation. The latest framework we're launching with a customer speaks to Awaab's Law — making sure we can map data based on what the guidance tells us, to understand where you might not be compliant and flag those comments and issues as quickly as possible so you don't miss anything. I can see some questions have come in, so I'll address those at the end — thank you, Millie.

Hopefully that gives you an overview of collecting, categorising and prioritising. Ultimately we want evidence-based decisions, but we've also seen this framework gives insight and CX teams a clearly mapped process for how you're listening and acting on customer feedback, which is incredibly important to the regulator. It should also give your teams confidence — a lot of CX delivery falls down when teams feel unconfident in what they're delivering.

You'll notice I'm stopping at "prioritise", halfway through the loop, and it creates a bunch more problems. Once you've got the right data and assets, how do you convince teams to act on it? How do you make sure it's working? How do we know these prioritised things are actually making a difference to tenant satisfaction? That's the next phase of work. Ideally we want to reduce the time you spend on the insight loop so you can spend more on CX delivery, project management and measuring impact. With that in mind, I'm going to pause there — and Sarah can share some bits on what we can offer to help with that side.

Sarah Wilson: Wow, that was brilliant, thank you so much, Helen. Absolutely fascinating — I've never heard the Pareto analysis described like that. I think that prioritise-the-frameworks part is the most exciting thing we've seen in quite a while at Wordnerds.

I've got a lovely segue given to me by the customer success team on Tuesday. They did a webinar all about "From Insight to Influence" — so the four, five, six that Helen referenced. We had Jenny from Granger; she's a customer, ex M&S and P&O, and she was describing exactly this process: once you've prioritised, where do you go from there, how do you get from insights to influence? How do you get buy-in for those insights and take stakeholders on the journey with you? I'll send this in the follow-up so you can watch — it's got Jenny's case study and a section from our CS Zoe on behavioural psychology and frameworks you can use.

If you look at the graph in the bottom left, you can see the meteoric rise in CX at Granger. When Jenny started it was very much doing what they'd always done before — no real clue what drivers were working, disconnected improvement initiatives, no idea if they were moving the needle. Jenny references a quote from the M&S CEO: "what got us here won't get us there." All the time the regulations are tightening and tenants are expecting more, so it's about adapting to be best-in-class.

Jenny was spending days and even weeks on the earlier parts of the framework, like categorisation. She's been able to switch that from days and weeks to minutes with some of the frameworks we've applied, which means now 60% of her time is spent on the later part of the loop — the initiatives and projects that actually move the needle. Because of that we've seen a marked improvement in their KPIs, and she's been heralded as a champion. She talks about doing workshops with stakeholders, meeting them where they are by sticking within a familiar framework, and inspiring curiosity in the business.

As a next step, if you're thinking "where do I go from here, where am I compared to other people?" — we've just launched a brand new maturity assessment. It's a few questions you can go through at your own pace; it takes about the time it takes to drink a coffee, about five minutes. You go through six different categories, get an overall maturity score, and we send you a personalised next-step report. You can review your performance breakdown by category — a nice way of taking stock of your progress.

If you're interested in what Wordnerds does specifically — we try to keep these as non-Wordnerds-focused as possible, we're really not very salesy, which Pete is always annoyed about — you've got my details there. I've got a very simple email address, so drop me a message or book a chat. I'd love to talk about where you're at, what other people are doing, and how we help you get to those next steps.

Now, questions — the juicy part. You've still got a couple of minutes. Oh, 15 minutes, right on time, Helen, so refreshing for us.

We've got a question from Millie: I've got a question about frameworks — what's the most common framework your clients use, and do you have any real-world examples? I'll give my social housing perspective, and Helen can come in with a couple more. In social housing, the number one is the TSM framework. We've got a built and trained theme bank that has 126 pre-trained themes around the main social housing TSM categories. It's got loads of other things too, so it can be flexible, but that's the most common framework, and it's a really easy way of getting all your surveys aligned in a way you can trust and getting the numbers Helen talked about.

In terms of real-world examples, I've got a lovely case study with Theo from Town and Country, who's a wonderful speaker. He talks about a specific example he found using the TSM framework — the cleaning of windows. People were complaining that workmen were using dirty water to clean the windows. He went into the data, put an exact number on it, and on who it was affecting and how. Because of that, the business installed new taps closer to the individual accommodations and used clean water for the window cleaning, which he then showed improved tenant satisfaction. That's a small nugget of how we go from big top-level categories right down to an actionable insight you can do something about. Helen, hit us with a retail or finance example.

Helen Precious: Obviously they don't use the TSM framework. Something useful that might relate to clients in social housing: it probably won't surprise you that customer feedback and employee feedback often feed into one CX strategy. You've got two perspectives either side of the table, and often when you need to find a solution, both stakeholders need to work together. So we do a lot with employee feedback and employee notes to understand how that might be impacting CX drivers.

Within that we use real research — things like the Hertzberg methodology. If you've done any kind of BTEC in business you're probably familiar with Maslow's hierarchy of needs; Hertzberg is similar, an iteration above Maslow in terms of understanding what employees need to feel motivated and successful at work. So we've got frameworks that look at that, and that feeds into a CX model, which is quite interesting.

Sarah Wilson: Thank you. Another one around the repair journey — I think one of our favourite frameworks is the repair journey. How does it work — do we need to capture the stages in the survey? Do you want to take this away?

Helen Precious: Yeah. The first thing to say about processes like this, where it's a journey, is not to think of them as linear, because often they're not. I'll direct you to the Insight to Influence webinar, where Jenny gives her example of how she mapped this practically with her team. But the principles, when you're looking at something like a repair journey and how you map the stages: firstly, get the business leads involved in the mapping stage, because their input is really useful.

The other thing is that, because it's not linear, think about all the touch points — things like what the quality of the repair was like, which doesn't necessarily fit into a flow where they started here and ended there. Some of this is more relationship-based, but it's equally important, and it does link, as we've seen in the data, to how people feel about the service they received. So, based on the data you've already got and a conversation with your leads, think about all the touch points, then map that as something you can present. We've got an interactive demo where we showcase what our repair journey looks like, so I can also share that if it would be helpful.

Sarah Wilson: Super helpful, thank you. We've got one from Linda — a really good one: from your experience, where does resident involvement tend to fit into frameworks? So in-person discussions, formal resident governance structures and resident groups. I'll give my perspective, Helen, then you can add to it.

There's a really fundamental change happening in social housing, where tenants are becoming much more empowered, and these involvement groups are becoming much more important — much more of a level playing field. Where these frameworks are super helpful is you can take the feedback you're getting and show the impact of what you're doing. So it's not just "you said, we did" — it's "you asked us to do this, and we haven't done that, and that's why." That tends to be really important for these conversations. It's about taking the feedback, knowledge and frameworks you've already got, and showing what you've done, how, and also why you've done it or why you haven't. What would you add, Helen?

Helen Precious: In terms of how we capture this conversation, I think there's probably work to be done as a sector on how we capture some of the in-person discussions and the feedback that goes into them. As soon as the resident has spoken about their issues or grievances in any format, they've told someone, so firstly, how do we capture all of this information? That really is in the collect stage of the framework.

The more important thing is on the impact side of the loop. The first stage there is designing a solution, and when we break down what "design" means, that's where the best people are co-designing. Give credit where it's due — social housing is far further ahead than a lot of other sectors when it comes to co-designing solutions with residents in the loop. This is why you need to spend 60% of your time on the impact side, on the delivery stage, because this stuff isn't easy. Co-designing a solution with your residents brought into that process is absolutely where it needs to start, and the best teams use working groups that bring employees and residents together to co-design solutions that work.

So largely this framework works to your question on the impact side: where can you bring them in, where can you test this information and make sure it works? Once you've designed a solution, is it the solution the residents wanted, does it work for everyone? That's where you need to start gathering more feedback and understanding what that solution looks like in real life. It's okay to test in one setting, but it needs to work for everyone. I hope that answers your question, and I'm more than happy to take feedback if you think we've missed the mark.

Sarah Wilson: We've got one last one — two minutes to go. For complaints, do we take only the initial complaint, or do you take the interactions between the provider and customer to resolution?

Helen Precious: Great question. We've done both, actually. We've looked at first point of contact as a driver of complaints — what the tenant said in that very first thing they log, whether in a repairs management system or over the telephone. We've also looked at the full relationship: the complaint and then all of the responses back and forth. We've looked at both of those interactions.

How we've done it is largely dependent on what you can get out of your systems in the first instance. Often we just export this stuff raw out of your CRM or complaints management system and take what we can get. We've also trialled — not in social housing — using an AI summary tool to summarise the issue from point one to resolution, just to reduce the duplication that comes through when we export it. This is getting into the detail of how the Wordnerds platform works: one of the things we do is look at the number of characters you put through the platform at its most granular level, so summarising helps with data volumes. Essentially we can do whatever we need to, and it depends on what we're trying to achieve with that complaint. I hope that's helpful, Robyn — happy to chat more if you've got a specific use case in mind.

Sarah Wilson: Robyn's from Home Group, so hopefully we can discuss in a bit more detail in the coming months.

Perfect. Well, I really enjoyed that. Thank you to Helen, who's battled lots of personal circumstances to be here — thank you so much for that brilliant section. I really hope you all enjoyed it and got some value from it. As always, we love your feedback, so you've got my email, drop me a message.

We're hoping to do our next one at the start of September, all around Awaab's Law and putting that framework into motion in a bit more detail, so stay tuned. And get in touch with me — if there's anything you'd like to see next, let me know, I'm always looking for content ideas. So thanks so much, go and get your lunch. That was a really helpful hour-long session. Bye everybody.