How long must a digital sign message hold? 8 seconds, usually

Digital signs cost more to run afoul of than static ones, in part because the rules governing them have an extra dimension: not just what the sign can say, but how long each message has to stay on screen before the display can change. Get the minimum display time — sometimes called dwell time or hold time — wrong, and a perfectly compliant-looking sign can still be in violation because it is cycling messages too fast. We pulled every minimum-display-time rule in our corpus to find out what the actual requirement looks like across the country.

The scale of the dataset

Our corpus contains 1,335 rows tagged with a minimum digital-sign display time requirement. Only 191 of them carry a machine-readable unit at all; the other 1,144 — 85.7% — state the requirement in prose that resists parsing, or state it only as a prohibition on “flashing” or “rapid” change with no interval attached. That is the same qualitative-drafting pattern we found in digital sign brightness limits, where 83.4% of brightness rules carry no unit either.

Of the 191 unit-bearing rows, 181 express an actual hold duration and 10 express something different — a maximum number of changes per minute, per hour or per day, which is the same idea inverted. Of the 181 durations, 162 are stated in seconds, 15 in minutes and 4 in hours.

The distribution, in seconds

A few things stand out:

• The middle of the corpus is tight. The 25th percentile (7 seconds) and the median (8 seconds) sit within a single second of each other, and 88 of the 161 numeric values — 54.7% — fall between 6 and 10 seconds inclusive. 43 of them sit at exactly 8 seconds, which is more than a quarter of the whole seconds-based population on one value. If you needed one number to design a default message-cycling schedule around, 8 seconds is the defensible middle-of-the-road choice.
• The low end allows genuinely fast cycling. The minimum value is 0.5 seconds — half a second — which is closer to a scrolling or animated display standard than a static-message-hold standard. Codes at this end of the distribution are permissive about how quickly content can change.
• The top quarter climbs fast. From the median (8 seconds) to the 75th percentile (15 seconds) nearly doubles, and the 90th percentile (30 seconds) is roughly quadruple the median. A sign programmed to the median 8-second hold could still violate a 90th-percentile code by cycling four times too fast.
• The slowest recorded requirement is a full 90 seconds — a minute and a half per message, which is a fundamentally different display strategy than the median 8-second code and would make rapid rotating advertising content essentially impossible to run compliantly.

The unit trap: minutes and hours are not typos

Alongside the 162 seconds-based rows, 15 codes state the hold in minutes — ranging from 1 minute to 30 — and 4 state it in hours, ranging from 1 hour to 24. None of these are extraction errors. A 30-minute hold is 1,800 seconds: twenty times the 90-second maximum of the entire seconds-based family, and 225 times the median. A 24-hour hold means one message per day.

This matters more than it might look at first glance. A code that requires a message to hold for “1 minute” is not a stricter version of an 8-second code — it is a code operating in an entirely different regulatory register, closer to how a static bulletin board might be governed than how a digital display board typically is. A technician who reads “1” in a code summary and assumes it means 1 second, when the ordinance actually specifies 1 minute, will under-hold every message by a factor of 60. Reading the unit is not a formality here — it is the single most consequential detail in the entire rule.

The 191 unit-bearing rows are written in 17 different unit spellings, including “seconds between changes,” “minute minimum interval,” “changes per minute” and “times per day.” Two of those describe a hold duration and two describe a change rate, and telling them apart requires reading the string rather than matching on it. That is a small instance of a much larger problem in this corpus: municipalities share no vocabulary, which is the subject of forty names for the same sign, and it is why we treat any count based on an exact string match as suspect until it reconciles against the total.

Why dwell time gets overlooked

Sign height, area, and setback get most of the attention in code-check conversations because they are visible at a glance during a site visit. Dwell time is different: it is invisible until an inspector times the display with a stopwatch, or until a complaint comes in about a sign that reads as distracting because it changes too fast. Digital sign controllers are typically programmed once at install and rarely revisited, which means a dwell-time violation can sit unnoticed for a long time — right up until it becomes the subject of a code enforcement action.

There is a second reason it gets missed, which the 85.7% no-unit figure makes plain: for most jurisdictions there is no number to look up. The ordinance says the sign must not “flash,” “blink,” or change “in a distracting manner,” and the enforceable meaning of that is whatever the zoning administrator decides it is on the day. In that situation the seconds-based distribution above is the most useful thing available — not because it binds you, but because a display cycling at the national median of 8 seconds is a far easier thing to defend at a hearing than one cycling at half a second. The rest of the electronic-message-centre rule set is covered in digital sign rules nationwide.

How to check the dwell time rule for your project

A report shows the minimum display time requirement for the applicable zone, including the actual unit the ordinance uses and whether it states a duration or a change rate, cited to the section — so you are programming the controller to the real requirement rather than an assumed one. Creating an account is free.