eCommerce Peak Season Cometh: Are you prepared to scale?

Peak season – a problem of scale

It’s the peak retail eCommerce season. Are you adequately prepared for scaling up to the deluge of new site visits and the heightened shopper demand for this period? It’s a known fact that shopper behavior changes significantly in peak season due to a combination of various factors: gifting, self-pampering, deal shopping, etc. As a retailer, you want to be prepared. In fact, lack of preparedness can mean a big difference between winning big or just cruising along.

1:1 personalization lets you scale

Gone are the days when marketers and merchandisers could get by by just using manual segmentation strategies targeting groups of online shoppers with the same products as a proxy for truly personalizing what they see. The reality today is that shopper behavior is changing so quickly that it’s impossible to scale with this manual targeting approach. To meet the demands of today’s hyper-competitive consumer demand, online retailers must deliver shopping experiences at an individual level across channels — at scale. Only an AI-based 1:1 personalization solution can achieve this.

1:1 personalization is a fundamental requirement for driving greater conversions and increasing shopping satisfaction, both of which lead to increased customer lifetime value and revenue. With 1:1 personalization on your side, online retailers are free from the tedious  and time consuming task of creating fragmented and localized merchandising promotion rules every week, which have been the bane of merchandisers. Instead, they can invest their precious time in producing great product descriptions and categorizations, choosing and testing variations, and letting the AI decide what’s best for the shopper based on their personal interests, previous interactions, and live clickstream. Effectively, 1:1 personalization can free retailers to focus on creating higher value for their eCommerce business.

In this post, we provide personalization tips and strategies to help prepare for the holiday season. These tips assume you are using an AI-powered personalization platform to scale your eCommerce business and achieve the best results.

Dig into your customer data

eCommerce peak season
Sales graph of outdoor apparel retailers with biggest sales peaks on are Black Friday, Saturday after Thanksgiving, Cyber Monday, Boxing Day, and Saturday after Christmas.

You should begin by analyzing historical sales data and identifying key trends.  Depending on industry segment, the number of peaks will vary in the holiday season. Within each of these peaks, it’s important to identify top product or category trends, determine  what customers are buying and for whom, and discern any changes in shopper behavior in order to formulate a game plan for maximizing revenue.

Sharpen your saw

Tune onsite search quality

Once you’ve collected  sales and customer engagement data, start by identifying the top searched keywords leading up to your peak season last year with low conversion rates. Here, the opportunity is to improve the conversions of these underperforming keywords by displaying more relevant results. Depending on actual visitor traffic, even a moderate improvement in the quality of onsite search results can significantly improve conversion rates and result in sizable revenue growth.

Using a 1:1 personalization platform, take the following practical steps for improving search results (as a best practice, be sure to incrementally test new personalization rules to see if they yield higher conversion rates):

  • Adjust relevance and ranking weights at the keyword level to improve the mix of products being displayed at a broad level.
  • Fine-tune keyword results by explicitly applying merchandising rules (boost / pin / bury / blacklist) to remove irrelevant products or low stock items, and push down low-performing or less relevant items.
  • Additionally, explore the potential of expanding the result set by creating keyword synonyms. These synonyms  provide more results for the searched keyword by associating it to other comparable keywords that better match the products for which visitors are searching.
  • Lastly, redirect high frequency search terms that don’t generate any search results towards existing category or landing pages to continue to engage the shoppers.

Increase organic traffic for top onsite search results

An additional key insight to glean from the historical customer engagement data is to identify high traffic keywords with high conversion rates, and create dedicated landing pages for campaigns such as Black Friday, doorbusters, flash sales, etc. Also, consider leveraging historical Google trends to supplement the list.

Specifically, seek to:

  • Create SEO-friendly landing pages for the identified keywords and ensure showcasing product results for the keyword. Reflektion can also automatically generate nested category pages (e.g. category > brand > color > size, etc.) based on product attributes for top product categories.
  • Include the SEO-friendly URLs of these landing or category pages into the sitemap for your eCommerce site. This placement ensures that search engine bots index the newly created landing page the next time they crawl the site.
  • Add personalized banners, navigation and taglines on the landing pages to improve visitor conversion based on a specific context.

By having SEO-friendly URLs for these landing pages, and by referencing them in your sitemap, you turn these high performing holiday keywords into search engine entryways to your site — increasing top of funnel traffic. With conversion rates holding steady, this tactic can  yield higher year-over-year revenues.

Improve revenue/visit by leveraging geographic trends

An often ignored aspect of personalization is that shopper choices — especially for apparel —  are shaped by the local weather, and to a higher degree, by what’s trending in the geo. For example, gifts that are popular vary greatly by geo where states such as Texas, Florida and Hawaii can be much warmer compared to northern states like Maine or Alaska — causing significantly different clothing preferences. Be it the holiday season or not, Reflektion’s AI automatically takes the geo into consideration by applying the contextual layer filter when presenting 1:1 personalized options.

However, Reflektion also provides advanced AI controls that allow merchandisers to take advantage of geo and plan their onsite (home page, landing pages, product detail pages or checkout pages) or email promotions — and win at scale. The idea is to leverage “collaborative filtering” (using AI to make automatic predictions about the interests of a user by collecting preferences or taste information from many users) to find correlations between the local holiday climate and your product categories to uncover merchandising opportunities and trends in major markets.


  • Reflektion allows eCommerce retailers to take advantage of prebuilt regionalized recommendation types or build their own automated regional recommendations or manual rules, especially if they are aware of strong correlations between product categories and regional weather. Merchandisers can build on this tactic by placing other recommendation types such as “Complete the Look” (for complementary items) or “You may also like” (for upselling related higher margin items) in the proximity of regionalized recommendations.
  • Merchandisers should ensure that these regional recommendations are compelling by a combination of testing for higher conversion rates and visual verification through shopper impersonation.

With the right geographically-powered AI-driven recommendations, merchandisers and marketers can use the power of collaborative filtering to scale their 1-1 personalization efforts to take advantage of real-time weather patterns, and local shopping trends auto-magically!

Leverage price sensitivity

Merchandisers spend a lot of time setting up promotional pricing during the holiday season to meet revenue targets. However, shopper behavior changes significantly (and continuously) during the holidays as they keep shopping for deals for themselves and on behalf of others. Constantly adjusting to their price sensitivity at scale is nearly impossible for merchandisers. But with a modern AI-powered personalization platform like Reflektion on their side, merchandisers can let the sophisticated AI engine learn the price sensitivity of the shoppers at the session level, and let it intelligently present products based on their learned price affinity in real-time and at scale.

If merchandisers wish to augment the AI, they can experiment with global rules and see if it has a positive revenue impact. This assumes that data on price sensitivity by region is available via Audiences in Reflektion Analytics or via Google Analytics. To be specific:

  • For price insensitive regions (where motivation to buy is high), merchandisers can boost more similarly or higher priced items and bury any sale or clearance items in recommendations or full page search results. This tactic not only improves the revenue per session but also protects the margins of the eCommerce retailer. For example: shoppers buying highly rated winter coats right before a major cold-snap.
  • On the other hand, for price sensitive regions (lower motivation to buy, or due to high competition), merchandisers can display more sale or clearance items to capitalize on those regional shopper behaviors. For example: buying Christmas decorations in areas with many vendors and very little product or price differentiation.

As a word of caution, this price-sensitivity approach needs to be tested a few weeks before every sales peak — and potentially in every major regional market — to be fully optimized.

Leveraging peak season learnings throughout the year

As you can see, the holiday season is a great time to test courageous merchandising marketing hypotheses that you may not have thought of otherwise. Interestingly, the prevalent perception in the market is that these strategies are only applicable to special times like the holiday season. However, upon closer review, the underlying principles are pervasive throughout the year. The reality is that these strategies and supporting tactics actually continue to work for the entire year! With that, here is our parting advice: don’t miss out on these year-round opportunities to take advantage of peak season best practices and never stop testing.

Frans Keylard

Frans is a data-driven digital marketing strategist at Reflektion and has 20 years of experience in enterprise eCommerce optimization and testing.