Cart recovery: retargeting ads versus email versus SMS, and what to use when

Cart recovery: retargeting ads versus email versus SMS, and what to use when is a decision point that can add thousands of rand to a monthly bottom line if handled with data‑driven discipline.

The real cost of abandoned carts

In South Africa the average e‑commerce conversion rate hovers around 2.5 %. That means for every 1,000 visitors, roughly 25 complete a purchase and the remaining 975 either browse or leave. Studies from the Baymard Institute show a global cart abandonment rate of around 69%, and South African stores commonly see figures in a similar range. If the average order value (AOV) is R1 200, those abandoned carts represent a potential loss of R1 200 × (0.71 × 1 000) ≈ R850 000 in revenue that never reaches the checkout.

Recovering even a fraction of that amount is worthwhile. A 10 % recovery rate on the same data set would generate R85 000 in incremental sales without any extra traffic acquisition cost. The challenge is to match the right recovery channel to the shopper’s behavior and the timing of the abandonment.

Channel comparison: retargeting ads, email and SMS

Each channel has a distinct reach, cost structure and response speed. Below is a concise matrix that highlights the key metrics you need to weigh.

Retargeting ads

• Reach: Up to 95 % of users who left a cart can be shown a display ad within the Google Display Network or Facebook Audience Network.
• Cost per mille (CPM): R12 - R25 in South Africa for standard display, higher for dynamic product ads.
• Typical click‑through rate (CTR): 0.2 % - 0.5 % for static ads, 0.8 % - 1.2 % for dynamic product ads that show the exact item.
• Conversion lift: 2 % - 4 % of clicks result in a purchase when the ad includes a time‑limited discount.
• Latency: Immediate impression after abandonment, but requires the user to be browsing other sites.

Email

• Open rate: 18 % - 28 % for cart recovery messages in the local market.
• Click‑through rate: 4 % - 7 % of opened emails lead to the checkout page.
• Revenue per email: R150 - R250 on average, depending on discount depth.
• Cost: R0.10 - R0.25 per email sent when using a bulk provider.
• Latency: Can be sent instantly, but the user may not read it for several hours.

SMS

• Open rate: 92 % - 98 % within minutes of receipt.
• Click‑through (link click) rate: 12 % - 18 % when the message includes a short URL.
• Average order value uplift: R200 - R300 per recovered cart, because the immediacy encourages a quick decision.
• Cost per SMS: R1.20 - R2.00 per message, depending on volume.
• Regulatory note: Must have explicit opt‑in; otherwise the message can be flagged as spam.

When you combine these numbers with your own AOV and profit margin, you can calculate a clear break‑even point for each channel. For example, if your profit margin is 30 % (R360 on a R1 200 order), a recovered sale via SMS that costs R1.50 per message yields a net profit of R358.50, making SMS the most profitable option for high‑value carts that were abandoned within the last two hours.

Step‑by‑step cart recovery workflow

Below is a repeatable process that lets you deploy the three channels in a logical sequence, ensuring you do not overspend on a channel that is unlikely to convert at that moment.

1. Capture abandonment data in real time

Integrate your checkout page with a tag manager (Google Tag Manager works well) and fire an event the moment the “Place Order” button is clicked but the transaction does not complete. Include these data points:

• Cart ID and product SKUs
• Cart value and any applied coupon code
• Customer identifier (email, phone number, or anonymous session ID)
• Timestamp of abandonment

Send the payload to a cloud function that stores the record in a database (e.g., AWS DynamoDB) and triggers the next steps.

2. Segment by time since abandonment

Define three buckets:

• Immediate (0‑30 minutes): Ideal for SMS and dynamic retargeting.
• Short‑term (30 minutes‑6 hours): Email with a gentle reminder and a small incentive.
• Long‑term (6‑24 hours): Retargeting ads with a stronger discount to re‑engage the shopper.

These windows are based on observed behavior: 40 % of recoverable carts are reclaimed within the first hour, another 30 % between one and six hours, and the remaining 30 % after a day of exposure.

3. Choose the channel and craft the message

Use the following decision tree:

• If the shopper provided a mobile number and opted in, send an SMS within 15 minutes. Keep the copy under 160 characters, include the product name, a clear call‑to‑action and a short link (e.g., bit.ly). Example: “Hi [FirstName], your [Product] is still waiting. Complete checkout now and get 10 % off: bit.ly/checkout123”.
• If only an email is available, schedule an email for 30 minutes after abandonment. Use a subject line that references the product, such as “Your [Product] is still in the cart”. Include a product image, a single‑click “Return to cart” button and a limited‑time discount (e.g., 5 % off for the next 2 hours).
• If the shopper is anonymous or the abandonment exceeds six hours, launch a dynamic retargeting ad. Feed the cart ID into the ad platform’s product catalog so the ad shows the exact items. Set a frequency cap of 3 impressions over 24 hours to avoid ad fatigue.

4. Measure and iterate

Track three key metrics for each channel:

• Recovery rate: recovered carts ÷ total carts in the bucket.
• Cost per recovered cart: total spend ÷ recovered carts.
• Profit per recovered cart: (AOV × margin) − cost per cart.

After a 30‑day test, compare the numbers. If SMS shows a recovery rate of 12 % at a cost of R1.80 per cart, while retargeting yields 3 % at R8 per cart, re‑allocate budget toward SMS for the immediate bucket.

A hypothetical worked example

Consider a concrete scenario to see how the numbers could play out.

Background

The retailer sells mid‑range apparel with an AOV of R1 500 and a gross margin of 35 % (R525 per order). In a typical week they record 2 000 abandoned carts.

Channel allocation

• 30 % of shoppers (600) have provided a mobile number and opted in.
• 50 % (1 000) have an email address but no phone.
• 20 % (400) are anonymous.

Implementation

They set up the workflow described above. The SMS message is sent at T+10 minutes, the email at T+30 minutes, and a retargeting ad campaign runs from T+6 hours to T+24 hours.

Results after 7 days

• SMS: 600 messages sent, 540 opened (90 %). 78 % clicked the link, and 66 % completed checkout. That equals 396 recovered carts. Cost: 600 × R1.50 = R900. Revenue: 396 × R1 500 = R594 000. Gross profit: 396 × R525 = R207 900. Net profit after SMS cost: R207 000.
• Email: 1 000 emails sent, 240 opened (24 %). 12 % of opens clicked, and 8 % of total emails resulted in a purchase. That equals 80 recovered carts. Cost: 1 000 × R0.20 = R200. Revenue: 80 × R1 500 = R120 000. Gross profit: 80 × R525 = R42 000. Net profit after email cost: R41 800.
• Retargeting ads: 400 anonymous users entered the ad pool. With a CPM of R20 and an average frequency of 2 impressions per user, total spend was (400 × 2 ÷ 1 000) × R20 = R16. Click‑through rate of 0.9 % generated 7 clicks, and 3 of those resulted in a purchase. That equals 3 recovered carts. Revenue: 3 × R1 500 = R4 500. Gross profit: 3 × R525 = R1 575. Net profit after ad spend: R1 559.

Overall, the retailer recovered 479 carts, turning a potential loss of 2 000 × R1 500 = R3 000 000 into an additional R594 000 + R120 000 + R4 500 = R718 500 in sales, with a net profit uplift of roughly R250 000 after channel costs. The data also shows that SMS delivers the highest ROI, email provides a modest lift at low cost, and retargeting ads are only worthwhile for high‑value or long‑tail carts.

To see where your own store is leaking, run a free scan of your store and get a baseline report.

Common mistakes and how to avoid them

Even seasoned operators can trip over a few pitfalls that erode the effectiveness of cart recovery efforts.

Sending too many messages

Bombarding a shopper with three SMSes in an hour leads to opt‑outs and brand damage. Stick to one SMS per abandonment, and if the first attempt fails, wait at least 4 hours before a retargeting ad.

Using generic copy

Personalisation matters. A message that mentions the exact product name and price outperforms a generic “Complete your purchase” by up to 30 %. Use dynamic variables from your cart data.

Ignoring mobile‑friendly design

Emails that are not responsive see a 40 % drop in click‑through rates on smartphones. Test every template with tools like Google’s Mobile-Friendly Test (Google Mobile-Friendly Test).

Failing to synchronize discounts

If you promise a 10 % discount in an SMS but the checkout code only applies 5 %, the shopper abandons again. Ensure the discount logic is triggered automatically when the cart ID is re‑opened.

Not respecting privacy regulations

South Africa’s POPIA requires explicit consent for SMS marketing. Keep a timestamped record of opt‑ins and provide an easy “STOP” link in every message.

Quick checklist for a high‑performing cart recovery system

• Implement real‑time abandonment tracking via a tag manager.
• Segment shoppers by time since abandonment and data availability (phone, email, anonymous).
• Send an SMS within 15 minutes to opted‑in mobile numbers.
• Send a personalized email at the 30‑minute mark with a modest discount.
• Launch dynamic retargeting ads after six hours for anonymous users.
• Use product‑specific copy and short, trackable URLs.
• Monitor recovery rate, cost per recovered cart and profit per cart for each channel.
• Adjust budgets weekly based on ROI calculations.
• Maintain compliance with POPIA and include easy opt‑out options.
• Run a free store audit regularly to spot new leakage points.