Peak season CRO: 6 advanced experiments to help you maximize profits during Black Friday and Cyber Monday

Kyle Hearnshaw

At the time of writing, there are seven weeks left until the Black Friday and Cyber Monday (BFCM) sales.

That means you’ve only got seven weeks left to make sure your website is turning visitors into customers as efficiently as possible before the floodgates open and the BCFM shoppers beat a path to your door (website).

But unfortunately, while many websites are generally quite well-optimized, very few have been optimized specifically for these high-traffic peak periods.

If this is you, then you could be leaving huge amounts of revenue and profit on the table during this year’s rush.

But have no fear: in this blog post we’re going to run through 6 of the most important tests that you can run to get your website in shape for the BFCM madness.

So, without further ado, let’s get into it.

Quick note: In the title and intro of this piece, we’ve emphasized the importance of testing if you want to make the most of the BFCM sales. Just to let you know: much (if not all) of the advice and many (if not all) of the test ideas to be laid out here are applicable to any kind of peak period – not just those surrounding xmas!

6 experiments to maximize profits during Black Friday, Cyber Monday, and other peak seasons

1. Sale promotion timing

One of the difficulties with any sale is knowing when to start promoting it.

If you advertise for Black Friday too early, you may prevent people from buying full-price products during the weeks leading up to it.

An example of an early Black Friday promotion taken from game.co.uk

An example of a previous early Black Friday promotion taken from game.co.uk

But if you advertise too late, awareness of your sale might be lower than you’d hoped.

Testing this can be a bit tricky, because you need a way to measure the value of advertising the sale to a user vs. not advertising it – and you also need to be able to do so without waiting until the sale itself.

This is where having something like an ‘early access’ campaign can help.

Rather than showing consumers a Black Friday advert and hoping they remember to come back, you can instead ask your users to sign up with an email address to receive early access to the sale (or, alternatively, to be notified when it starts).

Assuming you know the conversion rate and expected monetary value of your typical sale email campaign, you now have a measurable metric by which to judge the positive impact of the advert, i.e. no. of signups x average monetary value of a sale email.

Armed with this, you can periodically a/b test adding the Black Friday signup banner to your website in the months and weeks leading up to BFCM.

By comparing the negative impact of the signup banner (on things like full-price sales or conversion rate) with the expected value of early access signups, you can then work out when to start promoting the sale on your website.

Maybe a month out people aren’t in Black Friday mode yet, so you don’t get enough signups to justify taking up the space on the landing page.

But as Black Friday approaches, the potential marketing value of the signups might start to outweigh the conversion impact. This is the point when you should begin thinking about setting your sale promotion live on your site.

If you’re equipped for personalisation you can go one step further.

One thing we’ve tested in the past is only showing sale promotions to customers who we knew were discount driven. In this way you can ensure your premium consumers aren’t distracted by discounts.

        2. Sale countdown timers

Sale countdown timers can be an extremely effective way of creating urgency and guiding your users through to completing a purchase.

Unfortunately, over the years, consumers have become slightly desensitized – and even a bit skeptical – towards countdown timers.

That’s why it’s so important to a/b test your countdown timer execution to make sure it’s producing the intended effect.

For example, if you feature a prominent, flashing, second-by-second timer on your PDP, you may find that, rather than invoke urgency, you actually just irritate the user, which could deter them from following through with their purchase.

An example of a highly prominent countdown timer from asos.com

An example of a highly prominent countdown timer from asos.com

But on the other hand, if your timer is too small or discrete, it may not create enough urgency to actually produce any behavior change.

 

Example of a relatively discrete countdown timer – taken from Zaful.com’s previous website

We therefore recommend that you run some fairly extensive testing on your countdown timers to make sure you’re getting the most out of them. Here are a couple of potential tests you can run:

  1. Positioning of timer – should you have a sitewide banner with a timer or should you feature it only on specific pages in specific positions? If so, which positions? Which pages?
  2. When to start showing the timer – should you start the timer the second the sale begins or should you wait for the last couple of hours before it’s about to end? (if the timer starts too early, it may actually reduce the user’s sense of urgency!)
  3. Size and color of timer – should the timer take up half the page and be in bright red flashing colors or should it be a bit more discrete?

3. Promotion strategy: multiple discounts vs. flat discount %

A well-tested promotion strategy can make a huge difference during a high-traffic peak period.

For example, do you want to offer a flat discount of 10% on every item across your entire website or do you want to vary the discount you offer from one item to the next – maybe 15% off on all jeans, 20% off on t-shirts, etc.?

And if you do want to offer a flat discount rate across your entire website, what rate would be optimal to maximize profit?

In an ideal world, these are questions you’ll have already answered well in advance of the BFCM sales, with a set of trusted promotional tactics that you’ve tried and tested on promotions throughout the year.

But if you haven’t managed to do this, there’s still plenty you can be doing right now.

For example, you could run a mini-promotion on your website and vary the discount percentage you offer to different groups of users. In this way, you can closely measure the impact different discount rates have on the conversion rates of different products – and you can then use this information to inform your promotion strategy.

Experiment screenshots where we offer a blanket

Experiment concept where, for the control, you offer every user a flat 10% discount, and for the variation, you offer them a 10% discount plus an additional 5% off for all shoe purchases.

Another option for testing this would be to send out emails to your users offering them discount codes of varying amounts and then monitoring how these different discount amounts impact your conversion rates.

4. Promo merchandising

Another thing to consider with your BFCM sale is how you want to promote and implement it on your website.

There are lots of different ways that you can do this, so you’re going to want to test as many of them as possible to make sure that you get this right.

For example, do you want to run a single Black Friday banner across every page of your website that takes your users to a product listing page (PLP) with all of your discounted products? Or do you want to offer different wayfinding options up front, for example by giving consumers the option to jump straight into the sale within a single category?

What about letting consumers shop by their budget with PLPs like ‘items under £20’? Or by their discount behavior with wayfinding options for 20% off, 30% off, etc.?

Example of wayfinding option taken from matalan.co.uk

Example of a wayfinding approach organized around price – taken from matalan.co.uk

By running these different variants against each other, you should be able to zero in on the promo merchandising strategy that’s optimal for your website.

And unlike many of the other tests we’ve talked about so far in this blog, these types of experiments can actually work well during BFCM, so long as you have all of the different variants lined up beforehand and that you closely monitor your experiments once they’ve begun.

Multi-armed bandit or sequential testing are two tools that can help you make decisions on the fly as data starts to come in.

5. Delivery & collection

A number of research pieces suggest that things like free delivery and in-store collection are particularly powerful motivators during the BFCM period. For example,

Any competitive edge can make a huge difference during BFCM, so these different delivery options are well worth exploring.

If you already have a good understanding of how free delivery impacts your profitability, then there are various tests you can run now to optimize the way your website uses these levers.

For example, you could experiment on finding a balance between your discount percentage and delivery costs. Would your overall conversion rate and margin be better if you offered consumers only, say, 15% off instead of 20% but then offered free delivery?

To test this in a controlled way, you could create an email campaign before Black Friday where you offer one set of users a 20% discount and another set of users a 15% discount plus an additional ‘FREEDELIVERY’ promo code.

By comparing the purchasing behavior of these two groups, you’ll gain a strong understanding of how important free delivery is to your customers, and, crucially, how it might impact your conversion rate and profitability during Black Friday.

Note: You could also potentially run this kind of email campaign during the BFCM weekend. This will let you gather learnings about your user’s delivery preferences that you can use to inform the strategy of your next peak period.

6. Full-price content during promotions

While much of the discussion so far has been about your sale items and how you should promote them, you’ll also need to think about what you want to do with your non-sale items during BFCM.

Do you want to commit 100% of your energy to promoting the sale, or do you want to promote some of your higher value, full-price products too?

Example of a Black Friday homepage takeover on the Debenhams website

Example of a Black Friday homepage takeover on the Debenhams website

If you choose the latter approach, how prominent should your sale promotion be relative to your promotion of full-price products? And do you want to target certain segments more aggressively with your sale, e.g. discount-driven users, while focussing more on full-price goods with others, e.g. premium users who are generally less affected by price.

Having experiment-backed answers to these questions will help you wring every last drop of value out of your BFCM sale.

One test you can run – which will require a certain amount of personalization capability – involves varying the prominence of your BFCM promotion for your top 5% of premium users.

For example, rather than doing an entire homepage takeover, you could simply include a small promo banner at the top of your homepage while keeping the rest of the premium campaigns on the page live.

During Black Friday, you could then run this version of the homepage against the full scale Black Friday promotion and measure how the two perform against each other

You may find that the promotion distracts your premium users from purchasing full-price goods, in which case it would make sense to minimize the prominence of the sale for these users.

But on the other hand, you may find that your premium users actually also purchase the greatest number of goods in your sale as well – in which case, it would make sense to expose them to the same Black Friday promotion as everybody else.

Ultimately, you can never know the result of these kinds of experiments in advance – which is why it’s so important to run them!

Final thoughts

With so much at stake during the BFCM period, it’s extremely important that you get your promo strategy thoroughly tested and refined in time for the rush.

Hopefully, with the help of some of the ideas presented in this blog, you’re in a much better position to do that.

Now it’s up to you to begin running your own tests and working out the best way to tackle your upcoming peak period.

Good luck!

Join 10,000 other people who get our newsletter updates