Zero Panic Black Friday: Customer Communication Plan

This is the second half of our Zero Panic Black Friday communication plan. In part one, we covered the internal side: a tested roster of who to call when something breaks, a rule for when to move to the backup, and one standing meeting where the problem gets worked. That plan gets the right people fixing the problem. This one makes sure your customers know what’s happening while they do. The two are meant to be used together, and this plan runs on the roster, the call test, and the standing meeting you set up in part one.

The customer side needs its own plan, and the Cyber Monday outage from part one shows why. Merchants locked out of their Shopify admin spent most of the day with no idea what was wrong or when it would end. Many of them said afterward that they had been left in the dark while the outage was happening, and the lockout ran for much of Cyber Monday. The loudest complaint was about the silence, not the fault.

Your customers will react the same way to your store. They can forgive a problem. They have a much harder time forgiving not knowing. A customer communication plan answers one question before you need it: when something breaks, what do we tell our customers, and who sends it?

What silence costs you over the weekend

On a normal week, a customer who hits an error might try again tomorrow. From Thanksgiving through Cyber Monday, they have a dozen other tabs open with similar deals. Silence costs you more than the sale in front of you:

• Customers assume the worst. A broken checkout with no message looks like a closed store or a scam.
• Your support inbox floods. Hundreds of people ask the same question, and your team spends the day answering it one email at a time.
• Customers retry payments. Without a message, some will try their card three or four times, and you end up sorting out duplicate charges next week.
• Social media fills the gap. If you don’t explain what’s happening, someone else will, and they won’t be as accurate.

A short acknowledgment in the first 10 minutes, followed by a fuller update within 30, prevents most of this.

Two roles: one fixes, one talks

During an incident, there are two jobs that can’t be done well by the same person at the same time: fixing the problem and talking to customers about it. Split them, and name both on your roster from part one:

• The technical lead works the problem: the developer, the platform owner, or whoever owns the broken area on your roster. Their only job is finding and fixing the cause. They don’t write customer messages, answer social comments, or reply to support tickets.
• The communicator (the communications owner on your roster) handles everything customers see: the banner, social posts, emails, and the saved replies for the support team. They don’t troubleshoot. Their job is to keep customers informed so the technical lead can work without interruptions.
• The decision maker approves anything that commits the business: extending a sale, honoring a price, or offering a refund.
• Each role has a backup with the same logins and access.

The split pays off because of how each job breaks when it is shared. A developer in the middle of a fix who stops to write a social post loses focus, and the post usually comes out too technical or too late. A communicator who starts digging into server logs stops updating customers. Each person doing one job is faster than two people doing both.

The split also gives you one voice. When only the communicator posts, the banner, the social post, and the support replies all say the same thing. Mixed messages are worse than no message.

The communicator doesn’t need the technical details. They need four answers from the technical lead, in plain words:

1. What can customers not do right now?
2. Is it getting better, worse, or the same?
3. Is there anything customers should do, or avoid, such as not retrying a payment?
4. When should the next update go out?

The standing meeting from part one is where this handoff happens. Both people join, the communicator listens, asks those four questions, and posts. The technical lead keeps working and never has to repeat the story to anyone outside the meeting.

If you’re a small store and one person does both jobs, separate them in time instead: post the acknowledgment template, set a timer for 30 minutes, and then start troubleshooting. When the timer goes off, post the update and set it again. Two minutes on communication up front buys you half an hour of uninterrupted work.

Decide what triggers a message

Not every problem needs a public announcement. A slow admin page that customers never see doesn’t. A checkout that fails for one in three shoppers does. Agree on three levels now, so nobody debates it during an outage:

If you’re not sure which level it is, treat it as the higher one. An update you didn’t need costs very little. A silent hour on Cyber Monday costs a lot.

Pick channels that still work when the site doesn’t

The most common mistake is planning to post updates on the website that’s down. Your channels need to work independently of each other, so that one failure doesn’t take your voice with it.

Two things to check for each channel: who has the login, and whether the backup has it too. A banner tool only your developer can use is not a communication channel during an outage.

Add one rule: when to post and when to update

Write this rule at the top of the plan, next to the escalation rule from your internal roster:

1. Acknowledge within 10 minutes of confirming that customers are affected. One or two sentences: you know there’s a problem, what it affects, and that you’re working on it. No cause, no fix time. This is the message that stops customers from guessing.
2. First full update within 30 minutes. Now add what customers should do or avoid, such as “your cart is saved” or “please don’t retry your payment,” and when they’ll hear from you next.
3. Update every 60 minutes, or sooner if something changes. Even “still working on it” is an update. It tells customers you haven’t forgotten them.
4. All clear when it’s fixed, on every channel where you posted the problem. People who saw the first message need to see the last one.
5. Follow up the next day if customers lost something: a missed sale, a delayed order, a double charge.

The acknowledgment can only go out in 10 minutes if nobody has to approve it. Make it a pre-approved template the communicator can post on their own, on the fastest channels first: the site banner if the site is up, and social media. Approvals are for offers and promises, not for saying “we know.”

Level 3 problems, where customers have already bought and something is late, skip the 10-minute rule. Nobody is stuck at a checkout, so there’s nothing to announce publicly. Email the affected customers the same day with a new date, and update them again if that date moves.

The all clear is the step most stores forget. The banner stays up for two days, or the social post saying “we’re having issues” is still the latest one on Friday night.

Write the messages now

Nobody writes well with a broken checkout and a ringing phone. Write your messages in November, have the decision maker approve them once, and store them where the communicator and the backup can both reach them. During an incident, nothing needs approval: you fill in the blanks and send.

Start with these seven:

Acknowledgment, first 10 minutes (banner and social)

We’re aware that [checkout / payments / our website] isn’t working for some customers right now. We’re on it and will share more by [time].

Site banner, 30-minute update

Checkout is still not working for some customers. Your cart is saved, so there’s no need to start over. Please don’t retry your payment until we post an update. Next update by [time].

Social post, 30-minute update

Update: some customers still can’t check out on our site. Your cart is saved, and please don’t retry your payment for now. We’ll post the next update here by [time]. Thank you for your patience.

Payment problem

If your payment didn’t go through, please don’t try again yet. We’re fixing an issue with our payment provider and will post here as soon as it’s resolved. If you were charged more than once, we’ll make sure it’s refunded.

Discount code not working

We know the [CODE] discount isn’t applying for some customers. We’re fixing it now. If you already ordered without it, reply to your order confirmation and we’ll refund the difference.

Shipping delay, email to affected customers

Your order [number] is confirmed, but it will ship later than we promised. We expect it to leave our warehouse by [date]. We’re sorry for the delay, and we’ll email you the tracking number as soon as it ships.

All clear

Checkout is working again. Thank you for your patience. [If you’re extending the sale: Our Cyber Monday prices are extended until [time] for everyone.]

If you sell in more than one language, start with the language most of your customers use and get those templates right first. Then translate them into your next largest customer language, and so on, while there’s still time to have a native speaker check them. Translating a message under pressure is how mistakes get published.

What to say, and what to leave out

Every good incident message does three things: it says what’s wrong in plain words, it tells the customer what to do, and it says when they’ll hear from you next.

• Say what the customer sees, such as “checkout isn’t working.” Not “a 502 error from our payment gateway.”
• Tell them what to do: wait, don’t retry, their cart is saved. Don’t leave them to guess whether to try again.
• Give a time for the next update. Don’t promise a time for the fix that you can’t guarantee.
• Take ownership: “we’re fixing it.” Don’t blame a vendor or platform publicly, even when it’s their outage.
• Keep it short enough to read on a phone. Skip the long apology before the useful information.
• Use the same wording on every channel. Different details in the banner, the email, and the support replies make customers trust none of them.

The vendor point deserves a word. Even when the cause is your platform or payment provider, your customers bought from you. “We’re working with our payment provider to fix it” is honest without pointing fingers.

Give your support team the same message as a saved reply as soon as it’s posted. Customers who write in should get the exact words they saw on social media, not a fresh explanation from whoever happens to answer.

Decide what you’ll offer before you need to

When a sale breaks, customers often ask the same thing: “will I still get the price?” If you have to find the decision maker to answer that, the reply takes hours. Agree on your options now:

• Extend the sale by a set amount of time, such as 12 or 24 hours, if checkout was down during the promotion.
• Honor the price for customers who show they tried to buy during the outage.
• Refund the difference if a discount code failed and they paid full price.
• Refund duplicate charges right away, without the customer having to ask twice.
• Offer a small thank-you, such as free shipping or a discount on the next order, for longer delays.

Write down which options the communicator can use without asking, and which need the decision maker’s approval. That one decision can turn a two-hour wait into a five-minute reply.

Test it on November 23

You’re already making calls on Monday, November 23, four days before Black Friday. Add 30 minutes for the customer side. The communicator and the backup each run through it:

1. Log in to every channel from the phone they’ll have with them over the holiday weekend: social accounts, email platform, SMS tool, support desk, and the site banner.
2. Turn the banner on and off with a test message, on a staging site if you have one. If only your developer can do it, fix that now.
3. Open the templates from their phone. Confirm they can find them without a laptop.
4. Check two-factor logins. If a code goes to someone else’s phone or an old number, update it.
5. Load the saved replies into your support tool so the team can use them in one click.

The most common gap is a login. The communicator is ready to post, and the social account asks for a code sent to a former employee’s phone. Find that on November 23, not during the outage.

Your customer communication checklist

• A communicator and a technical lead are named for each incident, each with a backup who has the same access as the person they cover
• Three incident levels are agreed, with who hears about each and how fast
• At least two channels work even if the website is down
• The timing rule is written down: acknowledgment in 10 minutes, full update in 30, then every hour, all clear on every channel
• Seven message templates are written, approved, and stored where both people can reach them
• Templates are ready in your main customer language first, then translated into any other languages your customers use
• Saved replies are loaded in your support tool
• Make-it-right options are agreed, with who can approve each one
• Every login and the site banner were tested on November 23 from a phone

Silence is the real outage

You can’t promise that nothing will break between Thanksgiving and Cyber Monday. You can promise that your customers won’t be left guessing. A message written in November, acknowledged within 10 minutes, and closed with an all clear does more for trust than a perfect week ever could.

If you haven’t read part one, start there. The roster, the call test, and the standing meeting are what make this plan work.

If you’d like help building your communication plan, or a full Black Friday and Cyber Monday readiness review, get in touch with AraGrow. We work with store owners and their teams in English and Spanish.