Ask most retailers what their customers want and you’ll probably hear what their buyers want. On the Ecommerce Coffee Break podcast, Bizrate Insights VP of Client Insights Cheryl Zajac put a number on it. Roughly 98% of the people who come to your site leave without buying anything. Build your customer insights strictly from purchase data and you’ve built them from about 2% of your traffic, right as the other 98% shows up for the holiday season.

This late in the season, that gap gets expensive fast. Promotions, inventory, and staffing strategies are locked in, so you aren’t overhauling your holiday customer experience this late in the year. But there’s plenty of time for targeted improvements on the issues most likely to hit a breaking point when holiday volume hits.

Shoppers will call out a problem before it shows up in revenue, support tickets, or delivery complaints. A promotion that confuses people, a checkout flow that stalls on mobile, and unreliable delivery dates are the friction points seen by shoppers, not just buyers, that you can fix before they show up in your dashboard as a drop in revenue.

The signals are already there. The work left is finding the friction while it’s still small enough to fix.

Customer Feedback Early Warning Signals

By now holiday plans are mostly set, but you still have time to optimize or correct the top concerns. To set your focus, follow these simple guidelines: 

  1. Correct what’s cheap to correct
  2. Monitor what you can’t change yet
  3. Escalate quickly when something starts to slip. 

All three of those steps depend on noticing a problem early enough for the response to matter. These are known as leading indicators because they point to what’s about to happen. Customer feedback is frequently the earliest indicator retailers see, because shoppers describe a problem the moment they’re experiencing it.

Most of what sits on a holiday dashboard are lagging indicators. Revenue, conversion rate, NPS, and return rate are like the ghosts of Christmas past—they report on what has already happened. To help you look ahead, we’re going to explain how you can focus on leading indicators to prevent experiences that have the potential negative impact.


You can’t completely change your holiday plan this late. You can change how fast you find out it isn’t working.

The Customer Signal Stack

Holiday dashboards are built to report on what happened after a customer made their decision. Conversion tells you people didn’t buy, but it won’t tell you that site search returned nothing useful for “gifts under $50.”

Depending on your tech stack, you may be able to follow a non-buyer’s entire path, from the search they ran to the product page they landed on to the moment they left your site. But that data doesn’t tell you why they left—and the why is where you find the fix. That matters most for the 98% who didn’t buy. They’re the ones who hit the friction, and not one of them will turn up in a post-purchase survey.

Bizrate Insights’ full-journey feedback tools help retailers cross-reference dashboard results with the true cause. For example, seeing support volume go up 22% could easily be read as a staffing shortfall until someone on your team realizes the FAQ page doesn’t explain how to request delivery updates and more than half of contact reasons include “unclear delivery status”.  

The dashboard names each symptom accurately, but introduces room for error when you’re left to assume the cause of the problem every time.

Close that gap by monitoring and acting on three inputs:

  1. What customers say. These are direct signals that help you move fastest and include surveys, reviews, comments, contact reasons, chat transcripts, and abandonment feedback. “The promotion did not apply,” “delivery range was too long,” or “the return policy is confusing,” all tell you exactly what to fix.
  2. What customers do. These are behavioral indicators that help determine whether the feedback reflects a broader pattern. Examples include:
    • Search exits
    • Product-page abandonment
    • Add-to-cart declines
    • Checkout abandonment
    • Mobile conversion gaps
    • Increased use of customer service or self-service tools
    • Repeated visits before purchase*

*Salesforce data found that online traffic grew faster than sales did during the 2025 holiday season. More browsing can mean careful research or mounting hesitation, and the behavior alone won’t say which. 

  1. What the operation delivers. These operational measures identify whether you can keep the promises you’re making. Examples include: 
    • Inventory accuracy
    • Cancellation rate
    • Promotion error rate
    • Order processing time
    • On-time shipment and delivery
    • Carrier exception rate
    • Customer contact rate
    • Return and refund processing time

When two of the three of these signals move together, you know you have an actionable fix and not an isolated complaint. One shopper reporting a promo code failure is noise. That same complaint alongside rising abandonment at the promo field and a jump in discount-related support tickets is a problem with a tight deadline attached.

That’s the working definition of a holiday meltdown. It doesn’t require a site outage or a supply chain collapse, just an ordinary flaw that when combined with increased volume, urgency, and thin staffing turn into something costly.


Pick your three or four highest-risk journey points and confirm you can see all three inputs for each one. Where you only have behavior and operations, you’re watching symptoms without a diagnosis that customer verbatims provide. 

Four Signals Worth Your Attention Right Now

Signal 1: Shoppers can’t find what they came for.

Listen for: “I couldn’t find it in my size.” “Nothing came up when I searched.” “I wasn’t sure which one to get.” 

Holiday shoppers have a unique version of this since they’re buying for someone else. This time of year, site filters, comparison details, verified reviews, and in-depth sizing are more critical than usual.

Watch for: Zero-result search terms, exits from search result pages, product page exits, add-to-cart rate by category, and repeat visits that never convert.

What you can fix ahead of the holiday season: pull your top 50 search terms from the last two weeks and clear the zero-result ones, confirm gift-oriented filters actually work, and make sure your 20 highest-volume holiday SKUs answer sizing, material, and comparison questions on the page rather than in a support ticket.


Treat weak conversion in a product category as a discovery problem until the verbatims prove otherwise.

Signal 2: The promotion works for you, but not shoppers.

Listen for: “The code wouldn’t apply.” “The sale price never showed up.” “I didn’t know shipping would cost that much.”

Deloitte found that 58% of shoppers described holiday shopping as stressful in 2025. This means they’re less willing to search through multiple tabs or read fine print about whether items in their cart are excluded from sales. 

Watch for: Abandonment at the promo code field, redemption rates against forecast, discount-related contact volume, and average order value drifting off plan.

One Bizrate Insights client, Suits Outlets, sorts its review verbatims into two buckets: complaints about information and complaints about quality. The first group points to needing better product descriptions, while the second is shared with production teams. 

What you can fix ahead of the holiday season: QA every active code end-to-end on multiple devices and browsers, clearly list exclusions before checkout, and use consistent promotion language across ads, email, banners, and the website so it reads the same everywhere.


Test the offers you’ve already built instead of building new ones. Promotions fail on execution far more often than on strategy.

Signal 3: Mobile is where small problems have a big impact.

Listen for: “The page kept freezing.” “It wouldn’t let me enter my address.” “I gave up and used my laptop.” 

Adobe reported that 56.4% of U.S. online holiday spending happened on smartphones in 2025. That means mobile shoppers are worth monitoring closely. Very closely. If shoppers get so frustrated with your site experience that they would need to change devices or come back later, there’s a higher chance they buy somewhere else and never register as a complaint at all. 

Watch for: Mobile versus desktop for conversion, checkout completion by step, payment failure rate, and form error frequency. 

What you can fix ahead of the holiday season: Complete a purchase on your top three mobile devices and browsers, including variables like promo codes, payment types, and a logged-in, repeat customer versus a first-time shopper. A bug that only affects one mobile browser won’t show up in your overall conversion rate. It will show up in the orders you didn’t get. (That’s why Bizrate Insights’ analysis framework tells retailers to break down checkout metrics by device and customer type before diagnosing anything.)  


Segment by device now, or find the gap in your post-season recap.

Signal 4:You’re promising a delivery experience you can’t verify. 

Listen for: “Where is my order.” “Tracking hasn’t been updated in four days.” “It says delivered and it isn’t here.” If you offer buy online pick up in store (BOPIS), add cancelled-order complaints to that list, since those usually mean store inventory and your website aren’t aligned.

Availability tells you whether a customer can place the order, which is in-stock accuracy and cancellation rate. Fulfillment accuracy tells you whether you can deliver it when and how you said you would, which is promised-date changes, on-time shipment, carrier exceptions, and contacts per order. Trust breaks on the second one, and it breaks fast. Bringg found that 50% of shoppers stopped buying from a brand after a single bad delivery experience.

Filing “where is my order” volume as a customer service metric is a common mistake. It’s a fulfillment and communication signal that happens to arrive through the service queue, and staffing up to answer faster treats the symptom while the cause keeps producing both tickets and lost customers.

Give a closer look at BOPIS shoppers since roughly one in five online orders during the 2025 holiday season involved buy online, pick up in store, and climbed to about one in three in the final five days before Christmas. 

Last-minute demand concentrates on the capability most dependent on inventory accuracy, so stress-test pickup in October rather than in late December.

What you can fix ahead of the holiday season: Tighten cutoff messaging, turn on proactive status updates for any order past its promised ship date, set an escalation rule for carrier exceptions before a backlog forms, and reconcile store inventory against your site for the SKUs you expect to move through pickup.

Post-delivery feedback closes the loop, since what customers say about packaging, accuracy, and whether they’d buy again is the earliest read you’ll get on January’s return volume.


Every “where is my order” contact is a delivery promise that didn’t land. Categorize them by cause this month and you’ll know what to staff for in December.

Deciding What’s Worth Acting On: Leading vs. Lagging Signals

Plenty of signals will move between now and Cyber Week. Five questions separate the ones that deserve a response.

  1. Is it changing? Compare against a recent baseline and the same period last year, and rule out normal variation before calling it a trend.
  2. Is it concentrated? Break it out by device, customer type, category, promotion, fulfillment method, carrier, and region. Holiday problems rarely start everywhere at once.
  3. Is it severe? Frequency and intensity are different measurements. A few customers reporting a gift that arrived broken deserve more urgency than a longer list of mild packaging complaints. 
  4. Is it confirmed? Remember, one input is a hypothesis, while three inputs across what customers say, what customers do, and what your operation delivers are a confirmed and actionable issue. 
  5. Can somebody own it? A signal with no owner and no realistic fix before Cyber Week belongs on the January list. Say that out loud now rather than carrying it as an open item all season.

The fourth question is the one teams skip, and Zajac’s own example shows what it costs. Crutchfield, a Bizrate client for more than 20 years, saw a return intent spike right before the return window closed, concentrated in a single product category. The data read as a product defect, and the seemingly obvious next step was a QA investigation. The verbatims said something else. Customers weren’t reporting broken equipment or missing parts, they were describing uncertainty about whether they had installed it correctly. 

The fix was clearer installation instructions, and a team trusting the dashboard alone would have spent a quarter re-engineering a product that worked.

That case also settles the leading-versus-lagging question. Returns are normally a lagging measure, though defects in early-season orders work as leading indicators for the much larger wave still unshipped. Read them that way in early November rather than in January.


Sort your signals into leading and lagging indicators before the season starts. Teams that skip that step will waste time in December reacting to “ghosts of Christmas past” metrics that were already history when they arrived in the dashboard.

Find the Problem Before the Volume Does

None of this calls for a new VoC platform or a reorganized team. The next step is identifying the friction points that could have the most negative impact under increased holiday volume, monitoring them closely, and having prepared plans so you can quickly put them into action.

Every retailer has access to the same AI-powered dashboards this season, and those tools are great at surfacing patterns fast. What they don’t do is understand what a frustrated shopper meant, connect it to the warehouse change that happened the same week, or decide which of six problems to prioritize under tight deadlines and revenue on the line. That judgment belongs to people who know the business.

Start with one review this week. Pull customer verbatims from the last 30 days, set them beside your behavioral and operational data for the same period, and find the few places where all three agree something is wrong. Give each one an owner, a check-in cadence, and a deadline. Your team will feel confident going into the holiday season with the data backing their strategy and fixes outlined and ready to implement quickly. 

Season phaseSignals to prioritize
Now through OctoberSearch and navigation, product information, mobile usability, promotion clarity, inventory accuracy
Early November and Cyber WeekSite stability, checkout completion, payment and promo errors, conversion by device, service contact volume
Shipping-cutoff periodDelivery confidence, promised-date accuracy, order backlog, carrier exceptions, tracking questions
Final days before ChristmasPickup availability, store inventory accuracy, substitutions, cancellations, last-minute service demand
Post-holidayReturns, damage, refund cycle time, product expectation gaps, post-delivery satisfaction