Black Friday Tracking Before the Day

A week before the busiest weekend

Black Friday is a week away, on November 23, with Cyber Monday three days later. For many online stores it will be the busiest weekend of the year.

It will also produce the year's most useful data: which emails sold, which posts brought buyers, which ads paid for themselves.

That data only exists if the tracking works before the first visitor arrives. Anything broken on Friday stays broken in the reports for good.

So this week is for checking, while there is still time to fix what the checks find.

What to check this week

• Orders reaching the ecommerce reports with the right totals.
• Goals for sign-ups, forms and taps on the phone number.
• Payment pages that send buyers back without taking the credit.
• Campaign tags on every link in every email and post.
• Advertising conversions recorded in Google Ads and any social pixels.
• A send log of what went out, when, and to whom.

The third causes the most confusion later: a store that sends buyers to a payment service's pages can find that service credited with most of the weekend's sales.

One real order, start to finish

The simplest test runs through everything at once: place a small order on your own store, with a real card and a promotion code.

The next day, find that order in the ecommerce reports. Compare revenue, tax and shipping with the receipt, and note the source Analytics gave it.

Then reload the order confirmation page and look again later. Some setups record a sale every time that page loads, so one order appears twice.

Refund the order afterwards. The processing fee is the price of knowing.

The payment page that takes the credit

Many small stores send buyers to a payment service to pay, then bring them back to a thank-you page.

Analytics treats that return as a fresh arrival from the payment site and credits the sale to it, so the email or ad that brought the buyer loses the credit.

The fix is the referral exclusion list, found under Tracking Info in the property settings. Add the payment service's domain, and returning buyers keep their original source.

Check the payment settings too. If buyers are not returned to the store automatically, anyone who closes the tab after paying never reaches the thank-you page, and the sale goes unrecorded.

Goals for everything that is not a sale

Not every useful action on a sale weekend is a purchase. Sign-ups to the email list and taps on the phone number count too.

Record each as a goal. A tap on a phone number loads no new page, so it needs a small piece of code to record it as an event first.

Sign-up boxes added for a promotion often confirm inside a pop-up, so a goal waiting for a thank-you page never fires. Submit each one while watching the real-time conversions report.

Campaign tags on every link

Campaign tags are a few extra words added after a link's address, for example ?utm_source=newsletter&utm_medium=email. Analytics reads them to learn where each click came from.

Without tags, clicks from mail apps on phones usually arrive with no source and are filed as direct traffic. The newsletter that sold most can look as if it sold nothing.

Tag every link in every promotional email and social post. Google Ads tags its own clicks, provided auto-tagging is on and the accounts are linked.

Most email services can add tags for you. Check the setting is on, then click a link and watch the visit arrive in the real-time report.

A worked example

A small sock company that designs its own patterns sells online from a studio unit, and emails customers who signed up at markets and on the site.

Last year it sent three emails over the Black Friday weekend, posted daily, and ran a few Google Ads.

In December the owner looked at what had worked. The largest source of revenue was the payment service, and email seemed to have sold almost nothing.

Both were side effects of the setup: buyers returning from the payment pages were credited to them, and untagged email links, opened mostly on phones, arrived as direct visits.

This year the payment domain is excluded, every email link is tagged, and a test order with a code has been traced from email to report.

A send log is running too, so next December's figures can be read against what actually happened.

Names that stay consistent

Analytics treats Email and email as two different values, so a tag typed three ways becomes three rows in the report.

Agree the names once, in lowercase: email as the medium for every email, the list's name as the source, and one campaign name for the weekend, such as black-friday-2018.

Never tag links between your own pages. Analytics treats a click on one as the start of a fresh visit, and the visitor's real source is overwritten.

A log of what was sent when

Reports show what happened, not why. A one-page log does: each email, post, ad change and price change, with its date, time and audience.

Add the unplanned moments too: the hour the site was slow, the item that sold out on Saturday, the offer that was extended.

Analytics can also pin short notes, called annotations, to dates on its charts. One per send makes the weekend readable at a glance next year.

When the figures disagree

On Monday the store, Analytics and the advertising accounts will each report a different number of sales. That is normal.

Some visitors block tracking scripts, and browsers now limit some third-party tracking: Safari by default, Firefox for anyone who switches its new blocking on.

Refunds rarely reach Analytics, so its revenue runs higher than the money kept.

Use the store's own orders for money, and Analytics for proportions: which sources and campaigns brought what share of the sales.

The counter-case

A store expecting a handful of orders can keep this short: the order screen shows each buyer's code, and a separate code per channel does much of a tag's work.

And tracking does not make a sale. A week spent perfecting reports instead of the offer and the stock is a poor trade.

For any store expecting real volume, though: test one order, tag every link, and write down what you sent.

Before next Friday

1. Place one real order with a code.
2. Find it in the reports the next day.
3. Exclude the payment service from referrals.
4. Tag every email and social link.
5. Agree lowercase campaign names in advance.
6. Test each sign-up form in real time.
7. Start the send log today.

Step three rescues the most data. Without it, careful tagging can still end with the payment service credited for the sales.

Why some of those numbers will differ from browser to browser is covered in Firefox adds tracker blocking.

Frequently asked questions

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.