The Second Site and the Remote Half: Still One Desk

Monday, 08:10, at Tollgate Foods. The new distribution centre, forty minutes down the motorway, opened on Friday. The first call of the week is from its shift supervisor: "The label printer's dead and there's nobody here. Lorries at nine." Marcus, one of the three technicians, knows what is wrong. The printer is the same model as the ones at head office and it will be the ribbon. The spare ribbons are in a cupboard beside him, forty minutes from the printer. While he is finding his keys, a second message arrives from the operations manager. He has helpfully set up a chat channel called "DC Support" over the weekend so that the new site "has somewhere to ask". Marcus now has two queues, one cupboard, and a drive.

A second site is any place where staff work that is not where the desk sits. And the remote half is the part of the organisation that works from home for some or all of the week. They are the same design problem, support where the technician is not. And the remote half usually arrives first and more quietly, because home workers do not have a shift supervisor with lorries at nine. Both break every habit the desk built on proximity. That means the walk over, the drawer of spares, the face at the window, the knowledge of who sits where. This article is part of our Scaling Support Through Growth series. It is about redesigning those habits once so that the third site, and the fourth, are a copy rather than a crisis.

It covers why the second site is the hardest threshold. Then come the queue that stays one queue, coverage across sites, and spares in two places. Next it covers remote-first design, the local champion and the limits of the role, and the service catalogue per site. The site visit, measuring the second site, a short war story, and the multi-site support plan to copy follow, in that order.

Why the Second Site Is the Hardest Threshold

Every earlier threshold in this series added something to a desk that stayed in one place; the second site changes what a desk is. A desk of three in one building runs on proximity without knowing it. The technician walks to the printer, opens the cupboard, recognises the requester in the corridor, and knows that the finance team's row is by the window. None of those is written down, because none of them needed to be. On the day the second site opens, each becomes a forty-minute drive, a cupboard in the wrong building, a stranger's name on a ticket and a floor plan nobody has seen. The desk does not get worse. Its methods stop applying, which from the second site looks the same.

The remote half broke the same methods a year earlier and nobody noticed. A home worker with a dead laptop does not phone a shift supervisor. They struggle on with a phone and a sense that IT is for the office. That silence is the sign to watch. A desk whose home workers raise far fewer tickets per head than its office workers is not supporting them better; it is supporting them less. The quiet department test from The Seven Ways Support Quietly Fails applies to the remote half exactly as it applies to a building. The series on Supporting Remote Workers owns the remote half in full. This article takes from it the design idea that fixes the second site at the same time.

The Queue That Stays One Queue

The first rule of the second site is that it does not get its own queue. And the pressure to give it one will arrive before the furniture does. A site manager wants "somewhere for our people to ask". A chat channel appears. A technician who visited in the first week gets the site's requests in his personal mailbox. Within a month the desk has three doors into two queues and a report that covers one of them. The operations manager at Tollgate was not wrong to want the site to have somewhere to ask. He was wrong about where the somewhere should lead. And the fix is to point his channel at the queue rather than to close it.

One queue with a site on every ticket keeps the reports whole, the ownership visible and the backlog honest. Add a site field to the ticket. Set it from the requester's location where the directory knows it and ask for it on the form where it does not. Route by what the request is rather than by where it came from. That is the design in Routing by Department, Form and Mailbox. A label printer at the distribution centre and a label printer at head office are the same category. They are worked from the same knowledge base article, by whichever technician claims the ticket. The site field tells that technician whether a fix means a walk or a drive. A tool that brings every channel, the web form, the mailbox, chat and the rest, into one queue, as QuerySolv does, removes the argument about doors before it starts. What remains is the site field and the discipline of using it.

The second queue is also how a second site acquires a local favourite. If "DC Support" is a separate list that one technician reads, that technician becomes the site's help desk. In that case, everything follows: the hero, the leave week, the mailbox nobody else can see. One queue is the structural fix for the favourite. And the rest of the fix is the champion role below, which gives the site a person without giving it a technician.

Coverage Across Sites

Cover means a named person responsible for each site in each working hour. The second site needs it written down because proximity no longer supplies it. For a site forty minutes away with no technician on it, cover has two parts. One is a remote response, which is the same response the whole desk gives. The other is a visit, which is the part that needs a rule. Write the rule in ranges the desk can keep. Set a remote response within the normal target for the priority. Set a visit within a working day for anything that stops the site working and within the week for anything that does not. Add a scheduled visit on a fixed day regardless. The coverage calendar that holds all of this, leave, cover, on-call where it exists, is owned by The Coverage Calendar. And the second site is a column on it.

Shift patterns and time zones change the arithmetic, and they usually arrive with the second site rather than before it. A distribution centre that starts at six and a head office that starts at nine have three hours a day in which the desk is closed and the site is working. And the decision about those hours is the after-hours decision, owned by the article on coverage models for small teams. The remote half adds the same problem in a softer form, and Time Zones and Coverage owns it. The thing to decide here is only this: which hours each site is covered, by whom, and by what means. Write that down and tell the site. That way, the shift supervisor at six knows whether the dead printer is a phone call or a wait until nine. Silence at six is what builds the shadow channel by seven.

Spares in Two Places

A spare is only a spare if it is where the failure is, and Marcus's ribbon, forty minutes from the printer, was stock. The second site needs its own cupboard, locked, with a short list of what is in it. The list lives in the knowledge base with the site's name on it. That way, any technician, and the champion, can see what is there without driving. The contents are the site's own failure modes. At a distribution centre, they include label printer ribbons and a spare label printer, two scanners, a handful of the standard laptop's chargers. They also include a spare switch, network cables and a laptop ready to hand over. At an office, they are the standard laptop, a dock, two monitors, a headset, cables and chargers. The strategy for loaners and spares in general, how many and of what, is owned by The Loaner and Spares Strategy. The stock discipline is owned by Stock Management for a Small Desk. The second site adds a second row to both.

Check the cupboard on every scheduled visit and restock it from the ticket history. Read what was used from it, what was fetched from head office instead, and what the site asked for that was in neither. A quarterly look at the second question, what had to be driven over, is the cupboard's shopping list. For the remote half, the cupboard is a courier. A swap laptop is shipped on the day the fault is confirmed and the dead one shipped back. The practice is in Shipping and Swapping Hardware. Both the cupboard and the courier exist for the same reason. The desk cannot be in two places, and the spare can.

Remote-First Design

Remote-first design means writing every process for the requester the technician cannot walk to. The people the technician can walk to enjoy the same process from closer. Remote-first design is the single idea that fixes the second site and the remote half at once, and it is owned by Remote-First Support Design. Applied to the second site, it means the knowledge base article says what to do without assuming a technician is present. The laptop can be rebuilt without being plugged into anything at head office. The password reset works from anywhere. The first step in every hardware ticket is a remote look rather than a visit. A desk designed this way visits the second site to deliver and collect hardware and to do the things that need hands, and for nothing else.

Remote control is the tool this design rests on. It comes with a consent and logging discipline that is not optional at any site. The requester agrees to each session, the session is recorded in the ticket, and unattended access has a written policy. Consent Every Time owns the rule. This article only repeats that the second site is where desks are most tempted to skip it, because the alternative is a drive. The drive is cheaper than the trust. (It is also, on a Friday, slower than the trust, but that is not the argument.)

The Local Champion

A local champion is a non-IT person at the second site with a small, written role: eyes and hands for the desk, never a technician. They can read a screen to the technician on the phone and turn a device off and on when asked. They can open the spares cupboard and hand over the laptop that is ready. They can receive and return the courier's parcels, and tell the desk which end of the building the fault is at. They cannot hold administrative rights, reset anyone's password, keep a key to anything but the cupboard, or decide a priority. The role fits on one page. The page is agreed with the champion's own manager so that the time is recognised. The desk thanks them in public once a quarter, because the role is a favour and favours that go unthanked stop.

The champion is also the biggest risk the second site carries. A helpful person with a cupboard key becomes the site's help desk within a month if the desk lets them. The sign is the champion fixing things, which they will, and the fixes not becoming tickets, which they will not unless the desk makes it easy. The rule is that the champion never fixes without a ticket number. The desk gives them the fastest possible way to get one. Examples are a form with the site pre-filled, a chat message to the queue, a phone number that logs the call. The hallway-request discipline in The Hallway Request Handled Well applies to the champion exactly as it applies to a technician: help, log, hand back the number. A champion who logs is an extension of the desk. A champion who does not is a shadow desk with a nicer title.

The Service Catalogue Per Site

What the desk supports differs by site, and the difference should be written down before the site asks. Head office has meeting rooms, a reception display and a hundred laptops. The distribution centre has scanners, label printers, a ruggedised terminal on every forklift and a staff of forty who share four computers. The catalogue idea, a plain list of what the desk supports and how to ask, is owned by The Service Catalog Idea. The second site needs its own page of it. That page lists the devices and systems at that site and who supports each (the desk, the supplier, the facilities team). It lists the response the site can expect remotely and on a visit, and the one thing the site should never do. That last one is usually "unplug the switch in the cupboard to see if that helps".

Write the page with the site manager, not for them. Agree the response column in ranges the desk can keep rather than the ones the site would like. A site that has been told "same day if a lorry schedule depends on it, within the week otherwise" will accept the week. A site that has been told nothing will assume the drive is the default and will be disappointed by Wednesday.

The Site Visit

The scheduled visit is the second site's walk-up window: a fixed day, a fixed time, announced to the site, during which the technician does everything that needs hands. The visit list comes from the queue, filtered by the site field and by the tickets whose notes say "needs a visit". That way, the technician arrives with a plan rather than a greeting. Travel time is a ticket of its own, or a line on the visit ticket. That way, the hours the second site costs are visible when the capacity check in When to Add Heads is run. A desk that does not count the drives will find a gap it cannot explain.

Every visit attracts add-ons, because the technician is a rare visitor and the site has been saving things up. "While you're here" is a whole series, and Capturing Add-Ons as Tickets is the part that matters at a second site. Each add-on becomes a ticket on the spot, with the site field set. So the visit's real workload appears in the record and the next visit's list is longer than the queue alone would suggest. A visit that closes four tickets and fixes nine things has closed four tickets and lost five. Log the nine.

Measuring the Second Site

Measure the second site as a site, never as a person, with three numbers. Those are tickets per head per month compared with head office, response time by site, and the share of tickets that needed a visit. The first number is the quiet-site test; a site raising a third of what its headcount predicts has found another route, usually the champion or the favourite. The second shows whether the site is being served slower, which it will be at first and should not be by the second quarter. The third is the remote-first score. And it should fall over the year as the knowledge base and the cupboard catch up. The method for all three, built for the remote half and identical for a site, is in Measuring Remote Support. Put the site's row in the monthly report beside head office's, and expect the site manager to read it, because they will.

A Short War Story: The Two Queues at Bellweather Logistics

Bellweather Logistics opened a second depot an hour from its main site with a desk of three at the main site and no plan. Within a fortnight the depot manager had set up a chat channel for IT problems. By then, a technician called Dev, who had done the opening-week visit, had become the person the channel addressed by name. Dev answered everything, quickly and well, from his phone, and none of it became a ticket. The spares stayed at the main site, because nobody had asked for a cupboard. The depot's scanner tickets, all twenty-two of them in the first quarter, existed only in the chat. In the fourth month a scanner firmware fault took down eleven units in a morning. Dev was on leave. The other two technicians had never seen the channel. The depot picked by hand for a full shift while a technician drove the spares over. On arrival, the technician discovered that the fault needed a supplier call the chat history would have shown. The shift cost Bellweather a day's dispatch for a depot and an apology to its largest customer. The fix took a week. It was one queue with a site field, a cupboard with two charged scanners, and a one-page champion role for the depot's shift supervisor. The chat channel was pointed at the queue. A site with its own queue has its own help desk, and it is one person, and he is on leave.

The Multi-Site Support Plan

The plan below is one page per site, written before the site opens and reviewed on the first visit of each quarter. The top records the site and its people. The middle is the four fixes in this article, the queue, the cover, the cupboard and the champion, each with a name and a date. The bottom is the catalogue page and the measures. Fill one in for the remote half as well, treating it as a site with no building and a courier for a cupboard. Most of the plan applies unchanged, and the parts that do not are the parts the remote workers series owns. Keep the plans with the growth stage card, because the next site is a copy of this page with the names changed. That is the point of writing it down.

MULTI-SITE SUPPORT PLAN — ONE PAGE PER SITE
Site: ______________________   Opens / opened: ____________   Staff: ______   Hours: ________
Site manager: ______________   Reviewed on: ____________   Next review: ____________

1. THE QUEUE (one queue, a site field)
   [ ] site field on every ticket; set from the directory or asked on the form
   [ ] every door at this site leads to the queue: form ___  mailbox ___  chat ___  phone ___
   [ ] no separate list, channel or mailbox exists for this site (checked on: ________)

2. COVER
   Hours this site works: ________   Hours the desk covers it: ________
   Gap hours and what the site does in them: ______________________________
   Named cover by day:  Mon ______  Tue ______  Wed ______  Thu ______  Fri ______
   Remote response: normal target for the priority.   Visit: same day if ______________;
   within the week otherwise.   Scheduled visit: every ______ at ______

3. THE CUPBOARD
   Location and key holder: ______________________
   Contents (from this site's failure modes): ______________________________________
   Stock list in the knowledge base:  [ ]   Checked on every scheduled visit:  [ ]
   Fetched from head office last quarter (the shopping list): ______________________

4. THE CHAMPION
   Name: ______________   Agreed with their manager on: ____________
   Can: read a screen, power-cycle on request, open the cupboard, hand over, receive parcels
   Cannot: hold admin rights, reset passwords, decide priority, fix without a ticket number
   Fastest route to a ticket number for them: ______________________
   Thanked in public last on: ____________

5. THE CATALOGUE PAGE
   Devices and systems at this site, with who supports each: written  [ ]  agreed with site  [ ]
   The one thing the site must never do: ______________________________

6. MEASURES (in the monthly report, beside head office)
   Tickets per head per month: this site ______   head office ______
   Response time by site: ______   Share of tickets needing a visit: ______

Marcus wrote Tollgate's plan on the Tuesday, after the ribbon, with the shift supervisor as champion. Two ribbons, a spare printer and two charged scanners were in a locked cabinet by Friday. The "DC Support" channel stayed, pointed at the queue, with the site field set by the form. And the operations manager was told his idea had been kept, which was true. The drive still happens on Thursdays. It no longer happens at ten past eight.

Frequently Asked Questions