I noticed we were losing money. Our international orders had higher shipping costs, but we priced them the same as local ones. So I broke down all the costs by location. It was a tedious process, but it showed exactly which products were profitable and which were barely breaking even. My new rule is to regularly check the actual shipping costs for each region and then adjust pricing or bundle options for international customers. It's stopped us from throwing money away.
A crucial cost-to-serve revelation leading to an increase in margin was derived from customer segmentation based on order economics rather than revenue. By slicing the data according to gross profit minus fulfillment, returns, support tickets, and payment fees per order, a small group of high-volume, low-avg-order-value customers was identified as being consistently unprofitable, thus revealing a noteworthy trend. The analysis became clearer when we superimposed SKU-level return rates and shipping zones on it. The case of certain large, low-margin SKUs, again ordered individually, came into the picture. The action taken was not to let go of the customers but to redesign the model: We incorporated a minimum order limit for free shipping Merged low-margin SKUs into higher-AOV packs Established slower, lower-cost shipping alternatives for that segment.
We finally figured out why some clients were costing us money. A few specific B2B groups brought in good volume, but their regulatory hassle and support calls were eating our lunch. We adjusted their pricing and service tiers, which stabilized our margins and gave the team a break. The key was filtering by lifetime support cost, not just the initial sale price.
We broke down our service calls by job type and location and saw how much more some jobs were costing us. We were losing money on roofing and solar jobs in far-out areas because of travel and labor costs. So we adjusted our prices for those ZIP codes. Now, location-based pricing is our go-to when field costs get unpredictable. My advice is to figure out what your jobs actually cost in different areas. It helped us fix our profit problem.
I noticed some of our products were selling in tiny quantities but were a total pain to handle, basically losing us money. I threw the order frequency and handling costs into a simple scatter plot, and the problem items jumped right out. Seeing it all mapped out made it an easy call to cut a few things and set higher minimum orders. Honestly, sometimes the clearest answer comes from just looking at the data.
Here's something we found. Our oldest SKUs were eating up support hours like crazy, way more than any new products. It was a real money pit we didn't even see. We cut them and offered focused onboarding to the customers who used them most. The whole support process got simpler and our margins improved. I bet you have similar issues hiding in your data.
When we looked at our user data, we spotted a problem. Some subscribers were using a ton of expensive video content but were on our cheapest plans. So we changed things up. We pulled videos out of the standard package and sold them as add-ons. This evened out our revenue, and after some feedback, both our team and customers agreed it was a fairer way to handle pricing.
At Japantastic, some of our products were quietly killing our profits, specifically the bulky, fragile home decor items. I made a simple spreadsheet to break down shipping costs per product, and the numbers were eye-opening. We cut the worst offenders and raised prices on the rest. Our margins improved almost immediately. I suggest checking your costs per SKU regularly, because sometimes the smallest adjustments make the biggest difference.
Our biggest cost-to-serve insight was that revenue per client was a dangerously loose metric. We had some mid-market clients whose projects were profitable on-paper, but who eroded our margins every time. The signal was unmistakable as soon as we started tracking a high-level view of "non-billable delivery overhead": the ratio of project management and senior architect hours to total project revenue. This segment was three times higher than the overhead ratio of all our other clients, due to aggravating scope creep and high-touch demands. We were forced to re-design how we deliver service to this segment. Moving from a bespoke, high-flexibility model to a standardized, productized service with fixed scope and a clear menu of add-ons with fixed pricing made expectations crystal clear and lowered management overhead while actually stabilizing margin--without firing anyone. It forced upfront conversation about what's biologically possible for them within their budget, which wrote more predictable outcomes for everyone.
I tracked how much time my team actually spent supporting clients after launch and realized our flat monthly retainer was bleeding money. Small business clients with simple sites were emailing us constantly for tiny tweaks, while enterprise clients with complex sites barely contacted us. The data that made it obvious was logging every support ticket by client size and calculating hours spent versus revenue generated. Our $500/month clients were consuming three times more support hours than our $2000/month ones. I completely flipped our pricing to charge based on expected support volume, not website complexity. Added tiered support packages with defined response times and ticket limits. Margins improved by about 35% because we stopped giving unlimited access to our team's time for fixed prices.
We looked at our CLDY.com support data and noticed something weird. A handful of customers were generating most of our support tickets, but they weren't our biggest spenders. So we redesigned our service plans and added automated responses for simple questions. This freed up our team for our most valuable clients and our profits went up.
I realized that smaller custom orders required significantly more support time than larger ones. Segmenting customers by service intensity rather than order value made that clear. I adjusted pricing tiers to reflect actual effort, which improved margins without hurting satisfaction.
I grouped our service requests by region and noticed something odd. The city jobs were taking more of our time but weren't making us more money. So we changed our pricing, charging more for urban work. Within a month, our profit stabilized. It was a simple fix, but it showed me that looking at the specific cost data is the only way to get pricing right.
I looked at the numbers and couldn't believe it. Some of our bestsellers were actually losing money. So I sat down and figured out the real cost of each dish, ingredients and labor all in. We cut a few items and raised the price on a couple others by fifty cents. Suddenly we were making more money and the kitchen was less chaotic. Now I check item profitability all the time. It makes a huge difference.
President & CEO at Performance One Data Solutions (Division of Ross Group Inc)
Answered 2 months ago
Honestly, we crunched the numbers and saw our small clients, the ones needing constant hand-holding, were killing our margins. So we switched to tiered support plans where the service level matched their price tag. It took some careful conversations, but our support costs dropped and our team finally had time to focus on the big accounts that actually helped us grow.
I was looking at our numbers and something clicked. Our cheapest products were the ones getting all the customer service calls. The $15 widget needed way more hand-holding than our $200 kit. We tracked that for a month, then got rid of the worst offenders. It freed up our team to focus on the stuff that actually made money. Didn't fix everything overnight, but our ad spend started working better.
When I broke down our costs by property risk, I saw that urgent calls for high-risk locations were costing us money. So we created a new, more expensive priority service for those jobs. This immediately protected our profits and helped clients understand why a fast response costs more. I recommend you check your own service costs regularly to find similar hidden issues.
When we finally took a hard look at our product and customer profitability on a fully loaded basis, we realised we were making an implicit subsidy that eroded margins. We had always priced based on unit margin and average shipping cost, but we hadn't allocated the true cost of order handling, returns and customer support by segment. Using an activity-based costing model, we assigned warehouse touches, pick/pack labour, account management hours and transportation expenses to individual SKUs and customer segments. When we sliced the data by customer size and order pattern, we discovered that a handful of low-volume customers ordering our least expensive SKUs were consuming a disproportionate amount of fulfilment resources. The gross margin on those items looked healthy on paper, but once we incorporated the cost of processing frequent small orders and managing numerous returns, those SKUs were actually loss-making. That insight led us to redesign both our pricing and our service levels. We introduced a tiered fulfilment fee for orders below a certain threshold, encouraged those customers to consolidate orders through subscription or bulk programmes, and bundled the low-margin SKUs with higher-margin accessories to improve the overall order profitability. At the same time, we simplified our catalogue by discontinuing some slow-moving variants that added complexity without adding revenue. The methodology that made the signal pop was looking at contribution margin after cost-to-serve by cohort rather than by SKU alone. By sorting customers into quartiles based on order frequency, order size and return rate and then comparing contribution margin, we could clearly see where we were over-serving. That allowed us to have data-driven conversations with sales and operations about trade-offs, and ultimately lifted our gross margin by several percentage points without sacrificing customer satisfaction.
We worked with a founder who couldn't explain falling margins despite steady growth. Segmenting revenue by customer size didn't reveal much. Segmenting by onboarding hours per dollar of revenue did. The pattern showed small accounts draining the same support time as midsize ones while paying a fraction of the price. They redesigned pricing into two service levels. Self-serve for low-touch users, dedicated support for higher-value cohorts. The change cut cost-to-serve by almost a third within one quarter. The breakthrough came from pairing CRM activity logs with time-tracked support data. That cross-view exposed where growth was masking inefficiency and turned "more users" into "better users."
I run a marketing agency, and things changed when we looked at how much time each client actually took. We stopped taking on those small, high-maintenance projects. It felt weird at first, but our margins got better and the team wasn't so stretched. My advice is to track your service hours per client. You'll spot what to bundle and what to drop pretty fast. The real story is usually in the small stuff that eats your time.