The Complete Guide to Increasing Sales with Retail and E-Commerce A/B Testing

Inventory optimization can be challenging at times. Even seasoned Purchase Managers struggle to create informed inventory strategies. It may seem strange, but small changes can have a huge impact on what customers buy. For example, a liquor store may increase sales of a local whiskey if they display it on its own stand versus in the middle of an aisle shelf. If you’re not constantly optimizing your business for a higher conversion rate, you’re leaving money on the table.

Whether you realize it or not, you are regularly exposed to multivariate testing in your daily life. Large corporations, like Netflix and Amazon, use these techniques to increase engagement and measure success.

Large and small companies alike use A/B testing to inform inventory and business decisions that increase store sales. It’s simple to understand and easy to implement into any aspect of your business.

What is A/B Testing in Retail and E-Commerce?

At its core, A/B testing, or split testing, is simply trying out two different tactics to reach the same goal, like driving increased sales, then measuring the results.

A/B testing is common in e-commerce to increase ad conversion and website purchases. However, retail split testing can increase the in-store shopping experience by drawing more traffic, attention to specific products, and increasing average ticket values.

Why Should I be A/B Testing?

A/B testing can improve the customer’s buying experience leading to more sales. Happy customers mean a higher click-through conversion, increased revenue, and greater loyalty.

A/B testing pits product variations against each other to eliminate products or strategies that aren’t effective. That way you can maximize profits and conversion rates on the items you do sell. It’s not just about raising or lowering prices. It’s analyzing your inventory as a whole so you can reach your key performance indicator goals based on data, not intuition.

Shelf space is a precious commodity. A product should have to earn its spot in your inventory. If sales for one product go down, it may be time to A/B test it to see if it gets a spot on your next purchase order.

It also gives you the freedom to experiment. If you’re taking a risk on a different product brand, you can know whether that risk paid off in outperforming the inventory that was already there.

When it comes to inventory management, an ethos of split testing will ensure you are always in sync with your customers and your products are optimized to sell at the ideal price point.

More Benefits to A/B Testing

  • Easy, free way to improve sales, marketing, and inventory business goals

  • Turns ideas into measurable strategy

  • Informs inventory re-ordering decisions

  • Visibility on slow moving products that either need a new strategy or to be replaced by more profitable ones

  • Greater insight on how your business operates

  • Insight on what strategies work well with your customer base

  • Highlights products you should invest more in

  • Improve ROAS for e-commerce

When Should I Use an A/B Test?

Conducting a random split test is never a good idea. It inevitably means that you’re spending unnecessary time testing certain aspects that won’t translate to an increase in sales.

Before A/B testing, it’s essential to have a plan. This guide is intended to help you identify products or areas to split test, give examples of testing strategies, and how to get started and measure the results of your test.

What Should I A/B Test?

  • Slow moving products by physical store location or sales channel

  • Product and promotion verbiage

  • In-store layout and product merchandising

Retail A/B Testing

You can A/B test virtually every aspect of a physical retail store from product pricing to the selection of background music.

Recommended Retail A/B Tests

Product Placement: The way you display products in your store can substantially increase conversion. Try A/B testing the same product in two different locations, such as the checkout counter vs an aisle, and see if one drives more sales.

Product Options: Let’s say I own a liquor store and carry wine. I have a couple of upscale Pinot Noirs that sell for around $20, a couple mid-range bottles that sell close to $12, and bargain bottles for under $8. You can try removing the mid-range bottles and see if that leads to more upscale or bargain bottle purchases.

If more upscale bottles are sold you may consider removing the mid-range products if your profits increase. On the other hand, if more budget bottles sold it might be best for your profit margins to add the mid-range bottles back into your inventory.

Pricing: There are several ways you can A/B test pricing strategies. See how increasing the price for upscale options affects sales. Does it continue to sell just because it’s the ‘premium’ option?

Another A/B test is raising the price of a product by 10% to see if it increases profitability without affecting sales. Or you could try raising it to 20% and even though it might negatively affect sales, maybe the increased margins will more than make up the difference.

In-Store Signage: Track how different discount language impacts store traffic. One sale promote “20% off” then the next use “$20 off of $100 or more” to see what resonates better with your customers.

Opening Hours: Experiment with what hours your store is open to see how it affects sales, staffing schedules, and traffic.

E-Commerce A/B Testing

Online retailers use A/B testing constantly to optimize website leads and conversion rates. You can both split test your website’s product pages as well as any online ads, such as Google Shopping Ads, you’re running. A/B testing can bring to light significant optimizations for average order value, add to cart rates and conversion to checkout of site visitors.

While you can apply some of the pricing A/B tests explained above, you can also split test different ways to increase the online user experience.

A/B Testing for Product Landing Pages and Online Ads

  • Copy for the title, description, call to action button, or details

  • Leading product image

  • Social proof (testimonials, reviews, etc)

Tools for A/B Testing

The best thing about A/B testing is that you can do it for free as long as you have accurate sales data. This is where software comes into play. Whether you run a warehouse with 10,000 skus or a corner retail store with a few hundred items, data-driven decisions make businesses more profitable.

Reporting

Having robust reporting software, like Thrive, is a must for gathering and analyzing the data you need, affordably. Through A/B tests and a powerful reporting platform, you can get the right data in seconds, instead of hours, and make strategic decisions with confidence.

Security

By using A/B testing, retailers and e-commerce businesses can identify and reduce their risk of security breaches. With Evolve Security’s comprehensive services, businesses can uncover more vulnerabilities and benefit from expert guidance to ensure their product and customer data are secure. With the help of Evolve Security, businesses can take proactive steps to protect their customers and their bottom line.

Thrive's Robust Reporting Capabilities

Thrive gives you the data you need to make the best decisions on stocking and pricing. So you can let your competitors sell the stuff nobody wants at the wrong price at the wrong time of year, and in the wrong part of their stores.

E-Commerce Research Tools

Additionally, there are tools you can use to analyze your website or ad traffic and engagement. This data will help you identify areas that may be underperforming and what test results were most successful.

How is A/B Testing Done in a Physical Store?

  1. Identify products in your inventory that aren’t performing as well as you would like. Thrive’s Dead Inventory report is a great tool for keeping an eye on slow-moving products. For example, in a liquor store it may be a case of local beer.

  2. Designate your control factor and variable. This might be your current display (the control factor) versus adding a local sign (the variable).

  3. Begin A/B testing your control factor (testing your current display). Note, that if you already have accurate historical sales data for your control factor on a reporting platform, like Shopventory, you can skip right to the next step. Just make sure you’re comparing the same time periods.

  4. After you have completed your control factor test, begin testing your variable factor (adding a “Local” sign by the product). Ensure that you are running your test for the same amount of time as your control test.

  5. View your reporting to determine which test drove the most profit, traffic, or engagement.

  6. Make more informed inventory and business decisions based off data.

  7. A/B test additional ideas, one at a time.

Multi-Location A/B Testing

It’s important to realize that one strategy might not have the same results across all your stores. It’s likely that each of your locations has a slightly different customer base. For liquor stores, a specific brand of Chardonnay may drive a lot of profit at location A but may be slow moving at location B.

Shopventory’s powerful multi-location reporting capabilities mean users can view and compare sales performance for all their locations. These reports help business owners identify products at certain stores that could benefit from A/B testing.

After split testing, if a product at one location still isn’t succeeding, you may want to consider not stocking it at that store to make room for more profitable products.

How to Perform an A/B Test for an E-Commerce Website

Depending on your e-commerce platform and your level of confidence working on the backend of your website, there are two ways to implement A/B tests.

Client-Side A/B Testing

This method simply means making changes to your product page for all your visitors. For example, you may test image A for your whiskey for a month, then the next month try image B. You can make these split test changes on your e-commerce page editor.

Server-Side A/B Testing

Alternatively, you can use A/B testing software, like Optimizely, with your e-commerce platform.  This creates two different versions of your webpage (one version with image A and a separate version for image B). Then your web server will alter the different page variations to your audience.

Measuring Results for E-Commerce A/B Testing

You’ll measure the results of your online A/B test, same as you would in your store. Simply analyze sales and inventory reports to see what version moved the most products, was the most engaging, or contributed the most profit.

How to A/B Test Online Ads

Depending on how you’re running online ad campaigns, you can A/B test different variations of your ads to see which converts the best.

Google Ads

If you are managing your own Google Ads campaigns you can go about A/B testing one of two ways.

Manual Changes: Like website client-side testing, this approach lets you run an ad run with one variation for a specific amount of time before you make a change then run the ad again. It’s important that you let your ad run for at least 15 days before you make any changes so Google’s algorithm can optimize it.

Also, ensure when you are running your split test you are only making one change at a time, such as the product title.

Campaign Experiments: Before you launch your Google Ad campaign from your draft, you can set up an Experiment. Campaign Experiments use your same traffic and budget but allow you to split test different variations. To learn more about setting up a campaign Experiment, visit the Google Ads Help Center.

Advertising Agencies

If you work with a third-party advertising agency to run ads, simply connect with them to see what your options are for A/B testing.

How Long Should You Run An A/B Test?

We recommend running the control group and the variant test for a month each. That might mean testing a product’s regular price for one month, then testing a price increase for another month.

After gathering a month of data for each test variation, you can view sales reports for each timeframe to determine which method was more successful.

It’s also important to plan your tests around major holidays and shopping days so that data doesn’t sku your results. For example, it wouldn’t be a good idea for a liquor store to test wine prices over Thanksgiving because sales will be high that week regardless.

What Happens If My A/B Test Doesn’t Have A Clear Winner?

Let’s say you stock a bottle of cherry vodka that hasn’t moved in 3 months. You try A/B testing the location of the product in your store to make it more visible. If, after split testing locations, your sales data remains about the same, you may want to try another strategy, like pricing.

It is possible that after A/B testing a few different ideas that a product still won’t sell. If that is the case, it’s best to place a big discount on that product to get it out the door and make room for more profitable ones.

Power Up Your Retail Store

Learn more strategies and tools to grow your retail store, like cycle counting and inventory valuation reporting. 

Key Tips For A Successful A/B Test

  • Set clear goals for each A/B test that you can accurately report on using sales and inventory data.

  • Only run one test at a time per product. For example, don’t test a price change and a new image on an e-commerce product page at once. Then you won’t be able to attribute success to the appropriate strategy.

  • Ensure your testing is consistent.

  • Don’t assume that what works for one product applies to another. Different products can have unique target audiences or purposes. It’s best to run an A/B test first instead of guessing.

  • Having a stock forecasting tool, such as Thrive’s, will help you manage re-ordering since products may start moving at a different rate than you’re used to.

  • Never stop testing.

All Features Available During Your 30-Day Free Trial
All the tools you need to grow your business
PLAN DETAILS
STANDARD
Starting at $79/mo
PROFESSIONAL
Starting at $199/mo
ELITE
Starting at $399/mo
ENTERPRISE
Starting at $799/mo
Locations / Integrations
2
5
10
20
Users
2
5
10
20
Transactions Per Month
1-3,000
3,001-10,000
10,001-25,000
25,001-75,000
Products / Items / SKUs
Unlimited
Unlimited
Unlimited
Unlimited
Order History
3 Years
Unlimited
Unlimited
Unlimited
Reporting History
3 Years
Unlimited
Unlimited
Unlimited
Product Catalog Management
Catalog / Menu Management
SKU Creation
Lot Costs
Modifier Configuration
Pricing Calculator
Expiration Dates
Price Tag / Label Printing
Spreadsheet Bulk Upload
Average Costs
Inventory Management
Min / Max Levels
Low Stock Alerts
Stocktakes and Approvals
Basic Internal Transfers
Basic Purchase Orders
Modifiers Inventory Tracking
Kits / Bundles
Recipe / Ingredient Managment
Assemblies
Bill of Materials
Automated Purchase Orders
Automated Stock Transfers
Barcodes & Labels
Barcode Generation
Dymo Label Printing
Brother Label Printing
Avery Label Printing
Scan-to-Create Stocktakes
Scan-to-Create Invoices
Scan-to-Create Purchase Orders
Scan-to-Create Transfers
Reporting
Custom Dashboards
Sales Goal Tracking
Franchise Reporting
Product Profitability Reports
Inventory Change Reports
Dead Inventory Reports
Inventory Valuation Reports
Low Inventory Reports
Unsold Product Reports
Saved Reports
Region / District Reporting
Scheduled Reports
Intelligent Stock Forecasting
Wholesale
Customer Managment
Dynamic Price Lists
Wholesale Reporting
Payment Terms
Payments
Invoicing
Security
Theft / Discount Tracking
Multifactor Authentication
User Access Permissions
Support
Email
Live Chat
Phone
Dedicated Account Manager
Professional Services
Account Implementation
Get Quote
Get Quote
Get Quote
Get Quote
Catalog Migration
Get Quote
Get Quote
Get Quote
Get Quote
Product Creation
Get Quote
Get Quote
Get Quote
Get Quote
Extras
Global Currency Support
Dashboard Reporting App
Barcode Scanner App
Alexa Voice Command
Additional Transactions Available
Google for Retail Integration
QuickBooks Online Integration
Beta Feature Access