TECHNOLOGY · AGENTIC SHOPPING

From 6th to top 3: how Dell won in Alexa, Amazon's AI shopping assistant

In May 2026, Amazon folded Rufus into Alexa for Shopping, putting an AI assistant in the search bar, the app, and every Alexa device. Shoppers stopped typing keywords and started asking what to buy.

There are dozens of ways a product page can shape what Alexa recommends, and no obvious way to know which ones matter. Rather than optimise blind, Dell worked with WPP Media and Shalion to find out exactly where the brand was winning and losing.

 

9.7%→11.2%

Dell's Alexa share of voice, UK laptops, pre to post PDP update

 

+1.5 pp

Gained in a period when the overall market barely moved

 

6th → top 3

In Alexa share of voice for UK laptops

Action:

Scope: 48 weighted shopper questions, 16 consumer intents
Geography: UK
Category: Laptops
Time frame: Five snapshots, Dec 2025 to Apr 2026

 

The 5-step framework applied to Dell

ThThe same loop runs on any brand, category or market.

1

Query selection

Map the questions and consumer intents shoppers actually use to ask Alexa for a recommendation.

2

Monitor

Track share of voice by intent, against named competitors, trended over time.

3

Benchmark

See why a competitor is winning, whether it comes down to title, bullets or A+ content.

4

Diagnose

Match each intent to the SKUs in your catalogue best positioned to win it.

5

Act

Turn every content gap into a specific, copywriter-ready instruction, ranked by potential impact.

Result:

What the data reveals

“Partnering with Shalion during the beta test of their GEO tool was exciting and seamless. The granularity of the data stood out, enabling constructive, well-grounded recommendations for the test with Dell. Their eagerness to understand our needs made them feel like a true extension of our team.”
WPP-Media_Wordmark_Navy_RGB

Yara El Saadani

Senior Commerce Strategy Director - EMEA

Key takeaways you can apply to your brand

 

Be specific to win a recommendation.

Whatever gets implied, gets missed.

 

Specificity has a cost

If it is not managed deliberately and tracked, naming one use case can cost you another.

 

Reputation alone no longer earns a recommendation

Dell lost ground on the broad queries its old copy used to win by default.

 

This is diagnosable, not a guessing game

Every gain and every loss traces back to a specific question and a specific line of content.

See how your brand is winning and losing in Alexa's shopping recommendations.