top of page
2 days ago
3 min read

What should happen after a home improvement order ships?

By Trayan Hristov · Founder, Thrivelia · 19 November 2026

A bathroom cabinet can be delivered and still be weeks away from giving the customer any value. The installer may not be booked. One hinge may be missing. The cabinet could be sitting unopened on a building site. If “delivered” starts a seven-day review request and a matching-mirror offer, the brand is guessing about both.

Map the first usable outcome

For products that require installation, track order paid, dispatch, confirmed delivery, installation booked, installation complete and open issue. Some of those events belong to a carrier or installer. Record which ones you actually receive. If installation completion is unknown, ask one easy question such as “Is it fitted, or do you need a hand with anything?” Give the customer a support route that a person monitors.

A customer waiting for a fitter needs the correct model documents and a contact for scheduling. A customer reporting damage needs a case number and an owner. Suppress review requests and unrelated cross-sells while a case is open. When the case closes, confirm that the customer can use the product before restarting promotional messages. Use manufacturer-approved instructions or a qualified installer for installation questions.

Measure the gap between the warehouse and the home

Take an illustrative cohort of 100 delivered projects old enough for a 30-day check. Sixty are confirmed usable, 15 have open problems and 25 have no confirmed status. The observed completion floor is 60%. Reporting 60 out of the 75 known cases as an 80% completion rate conceals a quarter of the cohort. Keep unknowns on the scorecard and track the cost of finding out.

Break down the 15 cases. If nine concern the same absent fitting, the next action is with product, packing or fulfilment. A better sequence can route those cases quickly, but it can’t put a fitting in a box that has already left.

Change one handoff and price it

Pilot a delivery check on one product line. Compare the proportion confirmed usable by day 30, days to resolution, support contacts per project, refunds and contribution after service cost. Include the time spent making calls. A higher problem count in week one can mean your check found issues sooner. Follow resolution and repeat contacts before judging the pilot.

Only after confirmed completion would I test a complementary offer. Match accessories to the purchased model and actual finish. Keep a small group on service-only follow-up and look at net accessory contribution. For a different home-product context, The ClevHouse case shows how a product’s use can shape the next message. The installed-project example here is a proposed operating model.

Make the event flow recoverable

Installers sometimes close jobs in batches. A late completion event should close any outstanding “is it installed?” task, with one status update for the customer. Record a project ID, status, event time and source. If the same completion arrives twice, count it once. If a later damage report reopens the case, pause promotion again and keep the original installation date for cohort reporting. Test those two cases before connecting a service tool to automated messages.

Ask ChatGPT or Claude about your own numbers

Map these delivered projects into installed, waiting, unresolved and unknown. Which handoff causes the most delay or repeat contact? Define what should stop review requests and which team owns each exception.

Related reading

 
 
 

Comments


bottom of page