How to prevent abandoned carts: conversational ecommerce as prevention

A cart abandonment strategy doesn’t have to focus solely on recovering purchases after the fact. It can also act earlier, while a doubt is still in time to be resolved.

So before choosing a recovery sequence, it’s worth understanding why purchases get abandoned in the first place. In some cases, a timely answer can stop the customer from abandoning at all.

This article focuses on that earlier stage: identifying the doubts and stepping in while they can still shape the decision.

First: sort your abandonments

Not all abandonments have the same origin. Sorting them helps you identify which part of the process needs attention.

Take the last hundred and split them into three groups according to where the purchase broke down:

They left at the payment gateway. They reached the end, saw the total with shipping costs and didn’t continue. They may also have hit a mandatory sign-up or found their payment method missing. This group is mainly addressed from the checkout itself, by simplifying steps and showing costs up front.

They went back from the cart. They added the product and returned to the product page or the category. That behaviour can indicate they need to check some detail before completing the purchase.

They never reached the cart. They visited the product page, compared products and left. This group is harder to spot in your reports because it doesn’t always generate a clear purchase-intent signal.

The last two groups can be particularly interesting for conversational strategies. The first one mainly calls for checkout improvements.

If, once you’ve done this split, most of your abandonments happen at the payment gateway, your priority is optimising the checkout. We covered that in this other article, where we look at where the full funnel leaks.

The doubt that never becomes a question

A share of product doubts never turns into a query the store can see.

Someone who abandons over price has seen the total and made a decision. That behaviour usually leaves a clear trace: they reach the checkout, look at the amount and don’t continue.

Someone who abandons because they don’t know whether a lamp will fit their living room, whether a size M will work for them or whether a charger is compatible with their model may simply close the tab without ever starting a conversation.

The friction of finding a form, writing the question and waiting for an answer can be enough for that person to carry on their search in another store. And the window for resolving the doubt is limited.

The result is that your reports can show very few product queries even though there were plenty of doubts that never became conversations.

Stepping in earlier: the three moments that matter

Preventing abandonment means making help available at the points where doubts tend to appear.

1 · On the product page

This is where questions about sizes, measurements, materials, compatibility or real delivery times by postcode usually come up.

Some of these doubts can be resolved without a conversation. A visible, accessible size chart, for example, can head off queries and make the decision easier. It’s a good first move because it doesn’t require adding a new tool.

The doubts you can’t anticipate need a fast answer in the same place where they appear.

2 · When behaviour gives the doubt away

This is one of the differences between a passive chat and a conversational experience that accompanies the purchase.

A visitor who has spent several minutes on a product page, who compares the same two products repeatedly or who checks the returns policy before paying is giving you signals about what they need.

Triggers let you act on those signals and show a message related to the context. It might be an offer to help compare products, or a clarification about delivery times after they’ve checked shipping costs.

The underlying logic is called next best action: using real-time behaviour — what they look at, how long they take, where they go back to — to anticipate what information they might need next, instead of waiting for them to ask.

3 · In the cart, before they leave

The cart is another important moment for resolving doubts about when the order will arrive, what happens if the product doesn’t fit, or which payment methods are available.

These are objective questions that can be automated when the information is available. What matters is that the answer appears at the right moment.

Answering isn’t always enough: sometimes you have to show

There are doubts a written explanation may not settle, especially in highly visual catalogues.

“Do you have this model in blue?” — “Yes, in navy, sizes S to XL.” The information is correct, but showing the colour and comparing it against another option can help far more with the decision.

In categories like fashion, furniture, home decor or cosmetics, a visual answer can show the product in that colour, compare it with an alternative and let the customer add it to the cart from within the conversation itself.

It’s an experience similar to a salesperson in a physical store: as well as answering, they can show you the product to make choosing easier.

A visual answer can be especially useful when the choice depends on features the customer needs to compare.

Two channels done well beats five done halfway

When launching a conversational strategy, you don’t need to open every channel at once: web chat, WhatsApp, Instagram, Facebook Messenger or Telegram.

The priority should be the channels that concentrate your highest-value conversations and that your team can handle at the quality you expect.

A useful criterion is to look at where those conversations come from and concentrate your effort there. For many stores, the website and WhatsApp are good starting points, while the rest can be added as demand appears.

The channels you choose should share as much context as possible: customer history, catalogue and sales tools. If the experience differs sharply between the website and WhatsApp, the strategy loses part of its value.

When is this not worth it for you?

It’s also worth considering when a conversational strategy isn’t a priority.

If your catalogue is routine-purchase, low-ticket and comes with no attached doubts — consumables, standard spare parts or products the customer already knows — conversation may add less value. In those cases, simplifying the checkout can have a bigger impact.

If your visitor volume is small, you first need enough traffic for a strategy like this to produce a measurable return.

And if you don’t have the capacity to respond quickly or a clear automation strategy, an unattended chat can create a worse experience. It matters that the channel you open can be sustained at the quality it promises.

And if they abandon anyway?

Prevention doesn’t eliminate abandonment; it helps reduce it. For the cases that happen regardless, cart recovery still makes sense, and WhatsApp can be an interesting channel given its ability to generate opens and replies.

We already covered in detail how to approach that recovery message, and why it’s worth not opening with a discount, in this article. We won’t repeat it here.

What does add value from a prevention standpoint is connecting the recovery to the conversation that may have happened earlier.

Writing to someone who left without interacting is not the same as picking up a conversation in which, half an hour earlier, they asked about a size and were still deciding. In the second case there’s context, and the message can answer the specific doubt instead of being a generic reminder.

That’s why prevention and recovery can be part of the same strategy. A conversation started earlier can improve the relevance of the later contact.

And bear the cost in mind: WhatsApp charges per message sent. It isn’t a free channel, so that cost has to go into your calculation before planning a sequence at scale.

How do you know it’s working?

Compare your product page → cart ratio, alongside the overall abandonment rate. That’s the stretch directly related to the intervention, and it lets you measure its effect more accurately.

It’s also worth comparing the conversion rate of visitors who receive an intervention against those who browse without one. If there’s no difference, you may need to review the timing or the content of the message.

And keep an eye on the percentage of people who close the message immediately. A high figure can indicate the intervention is interrupting browsing rather than helping.

Give it a long enough period to gather data before drawing conclusions. If the product page → cart stretch doesn’t improve, the main opportunity may lie elsewhere in the process.

To close

Recovering a cart and preventing its abandonment are two distinct moments within the same strategy. Prevention lets you resolve a doubt while it can still influence the purchase decision.

The key is identifying those moments and offering relevant information without adding friction.

Oct8ne’s AI Agent works in that window: it identifies signals of interest or hesitation, steps in with information related to what the visitor is looking at, and can show the product inside the chat — on the website, WhatsApp, Instagram and Facebook Messenger, all from a single panel. If the query needs human attention, the conversation can pass to a person with the context intact.

You can request a demo and try it on your own catalogue.

And if, when you sort your hundred abandonments, you find most of them happen at the payment gateway, that’s useful information too: it tells you the priority is optimising the checkout before adding a conversational channel.