The downside of reliance on annual forecasting is that last year's data isn't going to be of much good on its own in 2020. But there's a huge benefit to the myriad interactions occurring online.

Angela Limoncelli, director of marketing for Luminoso

To say that projections around the 2020 holiday shopping season are fuzzy is being generously optimistic.

On the heels of the global pandemic upending supply chains and forcing nearly all interactions to digital channels, retailers and consumers alike are still reeling. For insights professionals tasked with advising retailers on projected customer behavior during these uncertain times, look to these three steps for guidance on how to navigate consumer data analysis coming holiday season.

Work with the data you’ve got

In our current climate, the downside of reliance on annual forecasting is that last year’s data isn’t going to be of much good on its own in 2020. But there’s a huge benefit to the myriad interactions occurring online: retailers can now track and interrogate most consumer touchpoints.

Most retailers will likely want a better understanding of what consumers like and don’t like about their new, involuntary ecommerce-only experience. And over the last six months, it’s likely many brands will have already collected feedback on these new experiences as they made the sudden shift. Mixed datasets – those containing both numerical and qualitative data, like NPS—will be of utmost importance. By having both ratings and contextual, conversational feedback, retailers can intimately understand how customers feel, about what and why.


With this in mind, start by looking at your brand or your retail clients’ questions over the last six months and evaluate which yielded the most insightful feedback. Based on these initial observations and analyses, consider what types of holiday-specific questions would be most useful for moving forward.

Define what success is—and isn’t—according to 2020’s rules

One thing it seems forecasters can all agree on about holiday 2020: consumers will value other things overspending on themselves. Although this might seem anxiety-inducing, it doesn’t mean consumers won’t be spending at all but will shift where they use their money. There are standard benchmarks, metrics, and KPIs retailers have used year over year to create projections and forecasts to meet and exceed. But attempting to improve on a narrative you’ve written in the past won’t work now.

Before holiday shopping begins in earnest, work to reframe expectations and projections based on the feedback you’ve been collecting throughout the past year. Whereas past predictions probably centered heavily on pure spending projections or craze around the hottest new item, used the nuanced data you’ve amassed to understand how and where you can maximize success:

  • Experience: If a product or offering attracts buyers based on emotions or holiday tradition, focus on metrics like CSAT and NPS. This data will feature nuanced feedback around experience and engagement.
  • Service: Is there a primary focus around solving issues as quickly as possible or providing market-leading service? Traditional service metrics, like CSAT and CES, are crucial to customer retention and those around resolution and response times. Leverage increased data from digital buying channels to learn more about how customers are newly engaging with service representatives.
  • New products or offerings: Even in 2020, brands will likely forge ahead with launches of the hottest products and services. Although historical data can help initial projections around purchase or adoption, go a step further and tap into understanding what consumers are currently thinking. Collecting and analyzing open-ended text feedback from forums, social media, and product reviews – in as near-real-time as possible during a launch – will be key.

Equip your team with the right tools

Whether you work for a research agency or in-house at a retailer, data and analysis quality will be critical to your success for both projections and a retrospective on a historic holiday season. A lot of your data is likely in the form of open-ended text: reviews, comments, survey responses, forums—the list goes on.


The good news is that there are text analytics applications available to help you and your team uncover the critical insights you need for success. Here are two key things to consider before investing in a tool:

  • No setup and ease of use: You, your analysts, and the team you’re delivering receiving the results you provide are likely working entirely remotely. If you need to put a new tool in place, steer clear of solutions that require technical knowledge and expertise to set up and use. Look for a no-code application that any analyst or business user can hop right into.
  • Quality of insights: The best way to maximize success is to understand what customers want. Informal language in reviews and forums is typically laden with jargon, slang, and misspellings. Look for solutions supported by powerful natural language understanding (NLU) with a native understanding of the languages you wish to analyze.

While there’s no way to know for sure what to expect in the 2020 holiday season, keeping these three themes in mind is a great start. A carefully-crafted strategy around data collection and analysis will bring focus to your team and holiday forecasting and reporting—and ensure you have powerful information to move forward for whatever’s next.

Luminoso provides text analytics applications for analyzing conversational text data like support tickets, open-ended survey responses and reviews.