PrefEngine
Preference driven product discovery

Channel Preference Is Not Content Preference, and Mixing Them Wastes Both

'Send me less email' and 'send me different email' are different requests. Preference systems that store one frequency number per customer flatten both, then wonder why unsubscribes stay high among people who asked for help.

Preference centres collect two different kinds of signal and most systems store them as one. Channel preference is about the pipe: email or SMS or post, how often, at what volume the customer starts to feel hunted. Content preference is about the payload: which categories, which product lines, which kinds of message, launches, replenishment, education, offers. A customer can love the content and hate the cadence, or tolerate daily messages in one category and want silence in another. When the system stores a single 'email frequency: low' flag, both customers get the same treatment, and both get it wrong. The category lover loses the launches they wanted because the volume cap throttles everything, and the cadence hater keeps getting the category they never wanted, just more slowly. Separating the two axes is not a sophistication. It is the minimum structure that lets a preference mean what the customer meant when they set it.

The two by two is where the money hides

Plot customers on the two axes and four groups appear. High content interest, low channel tolerance: your best audience for fewer, richer messages, and the group most damaged by a global frequency cap set from averages. High channel tolerance, narrow content interest: happy to hear often, furious when the topic wanders. Low on both: suppress with a clear conscience, they told you twice. High on both: the small group every program quietly depends on. Each cell wants a different program design, and a one dimensional preference gives all four the same design, which is why preference centres so often fail to move the unsubscribe rate they were built to move.

Related reading on this site: PrefEngine and Nobody Wants a Preference Center. They Want the Emails to Stop Being Wrong. cover the neighbouring problems in more depth.

Ask the questions separately or do not ask

The centre's form design is where the confusion is created. A single list of tick boxes mixing 'weekly digest', 'new launches', and 'SMS alerts' teaches the customer that these are the same kind of choice, and stores the answers in the same field. Lay the form out in two visibly different sections, how to reach you and what to reach you about, with the frequency question asked per channel, not per brand. SMS tolerance and email tolerance are different numbers in the same person, often by a factor of five, and a system that borrows the email answer for the SMS program will burn the SMS list's trust in a month.

Let behaviour arbitrate when the axes conflict

Stated channel and content preferences will sometimes fight the behavioural record: the customer who set content interest to one category and opens everything, or set frequency to daily and opens nothing. The working rule is that behaviour wins on channel, because fatigue is a fact the customer may not self report accurately, and the stated answer wins on content, because interest is a fact behaviour measures badly, since nobody can open a message about a category they were never sent. Apply that split and most preference conflicts resolve without a support ticket or a data science project.

Frequently asked questions

What is the difference between channel and content preference?

Channel preference covers how a customer wants to be reached: which channels, email, SMS, or post, and at what frequency on each. Content preference covers what they want to hear about: categories, product lines, and message types such as launches or replenishment reminders.

Why does a single frequency setting fail customers?

Because it flattens two independent choices into one number. A customer who wants fewer messages overall and a customer who wants the same volume but only on one topic receive identical treatment, and both end up dissatisfied for opposite reasons.

Should SMS and email share a frequency preference?

No. Tolerance for message volume differs sharply by channel in the same person; SMS is usually tolerated at a far lower frequency than email. Each channel needs its own frequency setting and its own fatigue measurement.

When stated preferences and behaviour disagree, which should win?

Behaviour is usually the better guide on channel questions, since fatigue shows in opens and clicks before a customer updates a form. Stated answers are usually better on content questions, because behaviour cannot reveal interest in topics the customer was never sent.

© 2026 PrefEngine · 447 Broadway, 2nd Floor #796, New York, NY 10013 · llms.txt