Enrolment Is the Whole Project: Getting Everyone Signed Up

Monday, 09:05, at Ashgrove Housing, six weeks after self-service reset went live. Priya runs the three-person desk, and the queue looks the way it looked in the spring: locked out, expired, locked out, "the new reset page says it doesn't know my phone". She opens the enrolment report, which she has been meaning to look at for a fortnight. Thirty-one percent of staff enrolled. Head office is at seventy. The housing officers who spend their days in tenants' kitchens are at nine. The repairs team, who work from vans and a shared tablet, is at four. The directory feature works perfectly. It has worked perfectly for six weeks, for a third of the company, and the other two thirds are on the phone. The project went live on time and has not started.

Enrolment is the step in which a user registers, in advance, the proof they will later use to reset their own password. It might be a phone number that receives a code, an authenticator app, a second e-mail address, a PIN, whichever the chosen approach uses. Without it, self-service cannot verify the person, so it cannot reset them. The page is a wall with a link to the desk at the bottom. A self-service system nobody has enrolled in is a reset ticket with extra steps. The enrolment number, not the go-live date, is the measure of whether the project exists. This article is part of our Password Reset Self-Service series, and it is about getting that number up and keeping it there.

It starts with why the enrolment number is the project, then the honest reasons people do not enrol and the fix for each. Next comes the campaign, as three messages with a reason in them, then the new-starter step that stops the number decaying. After that: the stragglers and how to reach them, the manager nudge, and the number to watch each week. Then come the forced-enrolment option and its cost, a short war story, and the enrolment campaign plan to copy.

A System Nobody Enrolled In

The mechanism is simple and it is why the number matters so much. The directory can only let a user reset their own password if it has a way to check that the user is who they say they are. The check depends on something registered beforehand: a phone, an app, an address. The registration is enrolment. A user who has not enrolled arrives at the reset page and is asked for the code that was sent to their phone. They have no phone registered, so they phone the desk. That is exactly what they did before the project, with one extra page in the way and a slightly worse mood. The feature was switched on for them. It cannot do anything for them.

So the reset count the project removes is capped by the enrolment number, and capped hard. If a third of staff are enrolled, the best the project can do is remove the resets that third would have raised. The enrolled third are usually the head-office staff who were already the easiest resets. So the removed tickets are the cheap ones, and the phone keeps ringing with the expensive ones. The reset-volume worksheet in Why Password Resets Dominate the Queue gave you a removable share. Multiply it by the enrolment rate and you have what the project is actually doing this month. Priya's project was doing about a third of what it had promised, and the promise is what the sponsor remembers.

This is why the project plan in The Self-Service Reset Project, Step by Step starts enrolment before the pilot rather than after go-live. It is also why every rollout wave has an enrolment target rather than a date. The technical switch-on is an afternoon. Enrolment is a campaign. That is to say, it is a piece of persuasion aimed at several hundred busy people who have not asked for anything. Persuasion takes longer than configuration. Nobody on a help desk was hired for persuasion. It turns out to be most of the job.

Why People Do Not Enrol

People do not enrol for five honest reasons, and none of them is laziness, which is worth saying before the desk starts composing reminders. They did not see the message, because it arrived in a mailbox they check weekly from a van. They saw it and did not understand what it was for. It said "SSPR" and "MFA", the industry's abbreviations for self-service password reset and multi-factor authentication, and "enrolment". None of those words told them that it meant fewer phone calls on a Monday. They understood and do not have the thing, because they have no work phone and were not told whether a personal one was allowed. They have the thing and do not trust it, because "give us your mobile number" from IT sounds like being tracked. Or they meant to, and it has been six weeks. They will do it the next time their password breaks, which is to say at the one moment they cannot.

Each reason has its own fix, and a campaign that only sends the message again is treating all five as the first one. The table below pairs them. The point of the table is that the fixes are mostly not messages: they are a different channel, a different sentence, a stated policy, a reassurance and a moment. The desk will know which reason dominates by department within a week of asking, and asking is the only way to find out. "Why haven't you enrolled?" asked kindly to five people in the repairs team will tell you more than the report ever will.

The Campaign: Three Messages and a Reason

The campaign is three messages in a fixed order, and the first one is not from the desk. The sponsor's e-mail goes first, short, in the sponsor's words. It says what is changing, why it is good for the reader, how long it takes, and by when. It also says that managers will be asked for their team's numbers. The reason has to be the reader's reason, not the desk's. Not "to reduce ticket volume" but "so that the next time your password expires on a Sunday night you can fix it yourself in two minutes instead of waiting for the desk to open". The desk's how-to goes second, linked from the first, a short article with a picture of each screen and nothing else. The reminder goes third, a week later, to the managers of the people who have not done it, with their names. That is the nudge below.

Promise the time honestly and keep it under three minutes, because the single most effective line in any enrolment message is a true small number. If enrolment takes two minutes, say two minutes. If it takes eight because the app has to be installed first, say eight. The campaign will lose some people but keep its credibility for the reminder. The message that says "quick and easy" and leads to a page that asks for an app, a code, a PIN and a security question will be remembered. The next thing the desk asks for will be met with that memory. The reader is being asked for a favour. Favours are granted to people who tell the truth about the size of them.

Put the link where the reader already is, not where the desk is. A link in an e-mail works for people who live in e-mail. For everyone else it goes on the intranet front page, or on the login screen if the approach allows it. Put it on a poster by the kettle or on the shared tablet's home screen. Bring it to the team meeting on the manager's laptop, with the manager doing it in front of everyone. The techniques for getting people to a page without nagging them into hostility are the subject of Promoting the Portal Without Nagging. Every one of them applies here. The campaign is marketing. The desk may as well admit it and do it properly.

The New-Starter Step

Every new starter enrols on their first day, during the same ten minutes in which they set their first password. Otherwise, the enrolment number will fall every month for as long as the organisation hires. This is the step campaigns forget, because a campaign is a push and a new starter arrives after the push has finished. Six months after a campaign that reached ninety percent, an organisation with normal turnover is back in the low eighties. The new people are the ones phoning on Mondays, and nobody can see why the number drifted. It drifted because enrolment was an event and hiring is a process.

Put it in three places. First, put it in the account-creation steps the desk follows. The account is created, the temporary password issued, and the new starter walked through enrolment on the first call or the first visit. All this happens before the welcome e-mail is sent. Second, put it in the manager's onboarding checklist, as one line, "enrolled in self-service reset, confirmed by the desk", so that the manager asks on day two. Third, if the approach supports it, put it in a prompt at first login that the new starter cannot dismiss. This is the one place forced enrolment is uncontroversial, because the person has no habit yet to be interrupted. A new starter will do anything on day one. By day thirty they are a straggler like everyone else.

Count new-starter enrolment separately, as a monthly line: starters this month, enrolled by the end of their first week. It should be at or near every one of them. When it is not, the gap is a process that has stopped happening. Usually the technician who did inductions has moved on, and the step lived in their head. The step that keeps the number up is the dullest step in the plan. It is the one that decides whether the project is still working a year from now.

The Stragglers

The stragglers are the people the campaign did not reach. They are nearly always the people whose resets cost the desk the most. That is the reason not to stop at two thirds. Field staff who log in from a van once a day. Shift workers in a depot whose e-mail is read by the supervisor on their behalf. Part-timers who are in on the two days the message was not sent. People on long leave, who will return with an expired password and no enrolment on the same morning. Senior people whose assistants manage their mailbox and would not forward "a thing from IT". And a quiet group who read the message, decided it did not apply to them, and were wrong.

Reach stragglers at the moment, not by the message, because they have proved the message does not reach them. The best moment is the reset call itself. A straggler phones the desk locked out, and the technician does the manual reset to the full standard. Then the technician says "while we are on the phone, let me get you enrolled so you can do this yourself next time". The technician does it, in two minutes, with the user's own device in their hand. This one habit turns every residual reset into an enrolment. A desk that does it consistently watches the straggler count fall by exactly the number of people who phone. It costs two minutes per call. It is the best two minutes in the whole project. Write it into the manual-reset script and check that the technicians are doing it. It is also the first thing dropped on a busy Monday, which is precisely when the stragglers phone.

The second moment is a visit. A technician takes a tablet to the depot for an hour at shift change, enrolling people one at a time as they clock in. This will reach more of the repairs team in a morning than six e-mails did in six weeks. The third is the team meeting, where the manager enrols on their own laptop while the team watches and then passes it round. The fourth is the return-from-leave conversation. The desk knows who is coming back because the leaver and absence list exists. It enrols them in the same call as the expired-password reset. Each of these is a few hours of a technician's time. Each is cheaper than the resets it removes within the month.

Separate the people who cannot enrol from the people who have not, and write the first group down. Consider the warehouse worker with no phone of any kind and the contractor who refuses a work app on a personal phone. Add the two accounts the security owner excluded, and the shared store login that has no single owner. These are not stragglers; they are exceptions. They get a named alternative route in the plan, usually the manual reset at the desk to the full standard. Counting them as stragglers inflates the gap and demoralises the campaign; counting them as exceptions makes the target honest. A target of "everyone except the fourteen we have listed" is reachable. A target of a hundred percent is a number that makes the person chasing it look like they have failed every week.

The Manager Nudge

The manager nudge is the sponsor sending each department's manager their own team's number, with the names of the unenrolled. This happens once a week until the target is met. It is the single most effective instrument in the campaign because it is the only one with any authority behind it. The desk cannot tell a sales team to spend two minutes on anything. A sales director can, and will, once a week, in the team meeting. That is if the number arrives in their inbox with their name on it and the department next door's number beside it. The report by department is a requirement of the approach for exactly this reason. A project that chose an approach without one will be building the report by hand from an export.

Write the nudge so that the manager knows what to do with it in ten seconds. Include their number, the company number, the target, the date, the names, and the link. Add one sentence in the manager's own interest: "the people on this list are the ones who will be phoning the desk from the car park next Monday morning instead of starting work". Do not shame anyone or publish a ranking to the whole company. Do not send the nudge to the unenrolled people themselves. They have already shown that messages from the desk do not reach them. The manager is the channel that reaches them. The desk's job is to make the manager's job take a minute.

The manager who replies "why should I care about this?" is asking a fair question and deserves an answer in their terms. That is the subject of Explaining Technical Things to Management. The answer is not ticket volume. It is their team's Monday morning. Think of the housing officer locked out in a tenant's kitchen with no way to look up the repair history. Think of the sales call that started twenty minutes late. Every manager has one of those stories. The nudge works when it reminds them of it.

The Number to Watch

The number to watch is enrolled-and-tested, by department, every week, against a target and a date. The word "tested" is the difference between a number and a hope. A user who registered a phone number and has never used it may have typed it wrongly. They may have registered a desk phone that cannot receive a text, or the old phone they replaced in the spring. Most approaches can tell you who has completed enrolment; fewer can tell you who has successfully used it. The pilot and the first wave should include a test reset for each enrolled person, and the campaign plan should say so. A number that says ninety percent enrolled and means seventy percent able to reset is a number that will embarrass the desk on the first bad Monday.

Plot it weekly beside the reset count, because the two lines are the project. Enrolment rises and resets fall. The gap between the removable share and the actual drop is the stragglers plus the exceptions plus whatever is wrong with the system. The reset count comes from the ticketing tool and the enrolment count from the directory. So somebody writes both into one sheet by hand each week. That is twenty minutes, and the most important twenty minutes of the week while the campaign runs. If you run QuerySolv, the volume reports give the reset line by month. The enrolment line will always come from the directory side, and the project is the distance between the two. The sheet itself, with the before and after columns, is in Measuring the Drop and Keeping It Down.

Expect a shape, and do not panic at it. The number climbs fast in the first fortnight, because the people who were going to enrol anyway enrol. It stalls somewhere around two thirds, which is where the campaign has reached everyone the message reaches. It then climbs slowly, one straggler at a time, through the reset-call habit, the visits and the nudges. It approaches the honest target over two or three months. Then it decays, slowly, unless the new-starter step is running, and the decay is the test of whether the step is. A desk that knows this shape in advance can tell the sponsor "we are at the stall, this is what the stall looks like, here is what we do now". That replaces explaining why the number stopped. Numbers that stop are not failures. Numbers that stop and nobody noticed are.

Forced Enrolment at Login

Forced enrolment is the option, where the approach supports it, of prompting the user at every login until they complete enrolment. There is no way to dismiss the prompt past a grace period. It gets the number up faster than anything else at the cost of a fortnight of mild resentment. It is the right answer for new starters, who have no habit to interrupt. It is often the right answer for the last stretch of a campaign that has stalled. This applies once the message has been sent, the managers have nudged and the visits have happened. It is the wrong first move. A prompt that appears before the sponsor's e-mail has explained it looks like something has gone wrong with the computer. The desk will get a ticket about the prompt from every person it was meant to help.

If you use it, announce it. Give a grace period of a week or two in which the prompt can be dismissed, and say the date the grace period ends. Have the no-phone alternative ready for the people the prompt will trap. Warn the desk about the week it goes hard. A prompt that cannot be dismissed on a shared store machine at opening time is a call to the desk at opening time. It will come from the store manager. The resentment passes. The resets do not, unless the number goes up, and the number goes up. Choose the fortnight.

A Short War Story: The Drivers at Bellweather Logistics

Bellweather Logistics was a haulage firm with four hundred staff, half of them drivers. A technician called Marcus switched on self-service reset and sent one all-staff e-mail. Marcus watched the enrolment number climb to forty percent in a month and stop. The drivers read their e-mail on a depot terminal once a week, if at all. The message had said "complete your SSPR enrolment" in the subject line, which to a driver is noise. For six months the reset count barely moved, because drivers were most of it. Every Monday the depot phones lit up with expired passwords for the delivery app. The desk reset them by hand, two hundred a month. Meanwhile, the feature that could have done it sat enrolled by the office staff who had never been the problem. The fix took three weeks. Marcus spent two mornings at each of the four depots at shift change with a tablet and enrolled drivers as they clocked in. Marcus told the desk to enrol every caller before ending the reset call. Enrolment reached ninety-one percent, and resets fell by more in the following month than in the six before it. The cost was six months of two hundred resets: twelve hundred manual resets that a fortnight of depot visits would have prevented. There was also a sponsor who had stopped asking about the project because it was "live". It had been live. It had not been enrolled.

The Enrolment Campaign Plan

The plan below is one page. It covers the target, the exceptions, and the three messages with their owners and dates. It also covers the new-starter step, the straggler moments, the weekly nudge and the weekly number. Fill in the exceptions list first, because the target is everyone minus that list and the list is what makes the target honest. The sample sponsor message is a starting point to be rewritten in the sponsor's voice; the only lines that must survive are the true time and the reader's reason. Review the plan weekly while the campaign runs and monthly afterwards, because the new-starter and straggler lines are permanent and the rest is not.

ENROLMENT CAMPAIGN PLAN — ONE PAGE
Organisation: ______________________   Campaign owner: ____________   Sponsor: ____________
Approach: ______________________   What the user registers: ______________________
Honest time to enrol: ______ minutes   Enrolment page reachable from: ______________________

TARGET
   Staff in scope: ______   Exceptions (cannot enrol, named route instead): ______
   Target enrolled-and-tested: ______ of ______ by ________
   Personal phones allowed for codes:  [ ] yes  [ ] no    No-phone alternative: ______________________

MESSAGES
   1. Sponsor e-mail       from: ____________   on: ________   [ ] sent
   2. Desk how-to article  owner: ____________   live by: ________   [ ] pictures of every screen
   3. Manager reminder     from: ____________   weekly from: ________ until target met

SAMPLE SPONSOR MESSAGE (rewrite in the sponsor's words)
   Subject: Reset your own password in ______ minutes, no call to IT
   From next ________ you can reset your own password or unlock your account
   from the login screen, any time, without waiting for the desk.
   To use it you need to register once. It takes ______ minutes: ____________.
   Please do it by ________. Your manager will have a list of who has.
   If you have no work phone: ______________________.
   Your number is used only to send you a code when you ask for a reset.

OTHER CHANNELS (tick what applies)
   [ ] Intranet front page   [ ] Login-screen link   [ ] Poster at: ____________
   [ ] Team meetings, manager enrols first   [ ] Shared-device home screen

NEW-STARTER STEP (permanent)
   [ ] In the desk's account-creation steps   [ ] On the manager's onboarding checklist
   [ ] Forced at first login where supported   Monthly check: starters ______ enrolled in week one ______

STRAGGLER MOMENTS (permanent)
   [ ] Every manual reset call ends with enrolment   [ ] Site visits planned: ____________ on ________
   [ ] Return-from-leave calls include enrolment     [ ] Enrolment clinic: ____________ on ________

WEEKLY, WHILE THE CAMPAIGN RUNS
   Enrolled-and-tested this week: ______ / ______   By department, lowest three: ____________
   Managers nudged with names: [ ]   Resets this week: ______   Sheet updated: [ ]

FORCED ENROLMENT (if used)
   Announced on: ________   Grace period ends: ________   Desk warned: [ ]   Alternative ready: [ ]

Priya's version of the plan at Ashgrove had fourteen exceptions, two depot visits and a line in the manual-reset script. The number went from thirty-one percent to eighty-eight in seven weeks. The repairs team was enrolled one van at a time in the yard. The housing officers were reached by their area managers in the fortnightly meeting, laptops open, in under five minutes each. The Monday queue at Ashgrove now has resets in it, a handful, from people on the exceptions list and new starters whose induction was missed. Priya reads the names each week to see which. The system had worked perfectly all along. It had just been waiting for someone to tell the repairs team it existed.

Frequently Asked Questions