Research
How can companies keep their promises as teams specialize?
How did companies discover customer problems, and who took responsibility for changing them? This research follows the records of what customers got and whether the approach lasted as organizations changed.
Research question
How are customer problems found, who changes them, and does the change last?
What is one problem for a customer becomes the work of several teams inside a company. Each team can do its part well while the customer still does not get the result they expected. I read papers, articles, and company records from the time to trace what companies tried, what changed for customers, and what the evidence leaves open. The current collection spans 1964–2026, from complaint analysis and consumer affairs to customer feedback programs and conversation analytics.
- Discovery · Could the company see the problem? How companies identified recurring causes in inquiries, complaints, and surveys, and which product, policy, or workflow needed to change.
- Action · Who decided to change it? When the team that found the problem was not the team that could fix it, who put the work on the agenda and why each team committed resources.
- Continuity · Did the approach last? Whether problems kept being resolved the same way after people or structures changed, and what was different when the approach stopped.
ROI is one consideration in allocating resources. The outcome I’m looking for is whether customers got what the company promised and whether teams could keep delivering it as the organization changed.
We Already Knew
Earlier attempts to turn customer problems into action
Companies have tried to connect customer problems to action for a long time. We Already Knew separates what was already known from what the records still leave unconfirmed.
-
Detecting recurring causes
What was tried · Namias proposed a statistical chart for identifying specific causes in complaint data.
What the record shows · The abstract describes the method. How firms acted on the findings is still open.
-
Bringing complaints into product discussions
What was tried · Texize’s consumer relations lead described working with executives, chemists, and sales staff.
What the record shows · The account documents reported collaboration. Product changes and customer outcomes still need corroboration.
-
Using surveys in product programs
What was tried · Caterpillar’s survey coordinator described using customer surveys to inform new product programs and assess quality indicators.
What the record shows · The trade article describes the process. Which changes followed and what they achieved remain open.
-
Following feedback through to action
What was tried · The authors examined how firms combined customer feedback, involved management, and tracked fixes.
What the record shows · The authors report gaps in their sample. Whether the same gaps persisted in other firms and periods is open.
-
Tracking contacts per order
What was tried · Amazon tracked the frequency of customer contact against order volume.
What the record shows · The shareholder letter reports improvement in that metric. Which customer problems became less frequent is a separate question.
-
Using conversation analysis to identify changes
What was tried · A vendor-hosted case describes how United brought contact patterns to teams that could make changes.
What the record shows · The account includes reported results. Independent evidence and durability remain open.
Reading the original sources
The first two analyses follow different records from 1980: a report on Texize and a model of complaint management. Both concern the step between discovery and action.
-
How did a complaints department take part in product discussions?
Reports how Nell Stewart described the changing role of consumer relations at Texize.
-
On what grounds did consumer affairs seek a voice in management decisions?
Examines how consumer affairs departments could gain influence within firms and how to evaluate their contribution.
Source archive
Publication details, original links, access and reading status, and related analyses.