The support budget is the smallest part of what customer contact costs a retailer. Here is how to build the number your finance team should actually be looking at.
Ask a retailer what a support contact costs and you will get the departmental budget divided by the contact count. Agent salaries, a share of management, the helpdesk licence, maybe the chat platform. For a mid-sized operation the answer usually lands somewhere between three and eight pounds per contact.
It is a real number and it is almost useless for decision-making, because it only captures the cost of answering. It says nothing about the cost of the thing that caused the customer to write in.
Work through the full picture and the same contact often costs several times the headline figure. That changes which problems are worth fixing and in what order.
Five layers, each one measurable from data you already hold.
Layer one: the handling cost. Fully loaded agent cost per minute multiplied by average handle time, plus the platform and management overhead. This is the number you already have. Keep it, but treat it as the floor rather than the answer.
Layer two: rework. A meaningful share of contacts are the second, third or fourth touch on the same underlying issue. Measure how many conversations relate to an issue that has already been contacted about, and you can calculate the true cost of resolving an issue rather than the cost of handling a message. In most operations the cost per issue is materially higher than the cost per contact, and the gap is pure waste.
Layer three: the resolution itself. Contacts do not just consume time, they trigger money. Goodwill credits, expedited reshipments, partial refunds offered to close a complaint quickly, return postage on items that were mis-described. Pull the value of financial gestures issued through support over a quarter and divide it by contacts in the same period. For many retailers this single layer exceeds the entire handling cost.
Layer four: the operational cause. A late delivery does not only generate a support contact. It generates a redelivery, sometimes a lost parcel claim, occasionally a return of an item that arrived after the customer needed it. If you can attribute contacts to root causes, you can attribute those downstream costs too.
Layer five: the revenue effect. The one nobody wants to estimate, and the largest. Customers who contact support and have a bad experience buy less afterwards. You can measure this without a model: take customers who contacted support in a period, split them by satisfaction outcome, and compare their spend over the following six months against a matched group who did not contact at all. The difference between a good support experience and a bad one, multiplied across the contact base, is usually the biggest figure on the page.
Once the full cost is visible, the priority order inverts.
If a contact only costs the handling time, the obvious move is to make handling cheaper: shorter calls, more automation, lower-cost agents. If a contact costs handling plus rework plus a goodwill gesture plus the operational failure behind it plus a dent in that customer's future spend, then the obvious move is to stop generating the contact.
That is a different project with a different owner. Delivery promises at checkout that the operation can actually meet. Accurate stock and product data so customers are not surprised on arrival. A returns process clear enough that nobody needs to ask how it works. Proactive notification when something has genuinely gone wrong, before the customer notices and writes in angry.
None of that is a customer service initiative. All of it reduces customer service cost, and improves the experience at the same time, which is the combination the cheaper-handling route can never deliver.
In retail, contact drivers are remarkably consistent. The largest single category is delivery status, some version of where is my order. Then returns and refund status. Then product questions that the site should have answered. Then payment and account issues. Then genuine complaints.
Look at that list as an operational scorecard rather than a support workload and each line becomes a question about a different team.
Delivery status volume is a measure of how much your customers trust your tracking. If the tracking page were accurate and the delivery estimate honest, most of those contacts would not exist. Returns status volume measures how visible your reverse logistics are to the customer. Product question volume measures the quality of your product data. Payment issue volume measures checkout friction.
The support team cannot fix any of those. They can only report them, which is why the contact-driver analysis belongs in the operations review rather than the CX one.
Take an operation handling a thousand contacts a day. At five pounds of handling cost, that is roughly one and a quarter million pounds a year. The number everyone quotes.
Now add the layers. Suppose a fifth of contacts are repeat touches on an existing issue, which means the cost per resolved issue is around twenty five per cent higher than the cost per contact. Suppose financial gestures issued through support average two pounds per contact across the year. Suppose the customers who have a poor support experience reduce their spend measurably over the following six months, and that group is a meaningful minority of contacts.
The fully loaded figure lands several times higher than the budget line, and the largest components are not staff costs. They are the money handed out to close problems and the revenue lost from customers who stop trusting you.
You do not need precision here. You need the order of magnitude, because it tells you where to spend the next quarter. Halving the handling cost of delivery status contacts saves a slice of layer one. Removing the reason those contacts happen saves layers one through five at once.
Take it to the operations and trading review, not the CX review. Present it as a list of causes with a cost against each one, and the split between what support can influence and what it cannot.
Then pick the largest cause that another team can fix, help them build the case, and measure the contact volume for that intent afterwards. A visible reduction in a whole category of contact is the most persuasive artefact you can produce, because it shows the same or better customer experience alongside a lower cost to serve, which is the only version of cost reduction that holds.
We join ticket data to orders, deliveries, returns and refunds, and come back with contact drivers costed line by line.