Klaviyo Flows Beyond the Templates: 7 Automations That Carry Weight

Almost every shop has the Klaviyo templates switched on. And almost every shop therefore sends the same three emails as everybody else.

The templates are a good start. They are just that: a start. With nano, mate and MUSTAX we run three of our own brands with two people. Email is one of the few channels that keeps working at night. So we have taken the flows apart and rebuilt them over the years.

Here are seven flows that really carry weight for us. Each one with trigger, content and purpose. And after that, the list of flows you can save yourself.

Why templates alone are not enough

The flows that ship with the tool do not know your business. They do not know whether your product is a consumable or a one-off purchase. They do not know when your parcel arrives. They do not know what question your customers ask on day three.

Those three things decide whether an email helps or annoys. At nano, the review request only goes out once tracking says the parcel has been delivered. That is a tiny condition, and it changed the response rate considerably. Before that, the email ran after a fixed number of days, so sometimes before the customer even had the product in hand.

The rest of this article is at its core a collection of conditions like that one.

Flow 1: Welcome series with a branch

Trigger: New newsletter signup.

Content: Three to five emails. First email immediately, with the incentive you promised. Second email after 24 hours with the story behind the product. Third email after 72 hours with the concrete use case.

Purpose: Turn a contact into a first-time buyer while the interest is fresh.

The difference from the template is the branch. As soon as somebody buys, they leave the series. That sounds obvious and often is not in practice: we regularly see shops chasing a fresh buyer with the welcome discount for another three days.

Second branch: somebody arriving from a product page gets different emails than somebody arriving from a blog article. One is close to buying, the other is doing research.

Flow 2: Abandoned cart in three stages

Trigger: Cart created, purchase not completed.

Content: First email after about an hour, purely service-oriented. Second email after 24 hours addressing the objection you hear most often. Third email after 72 hours with social proof.

Purpose: Rescue purchases that failed over a detail.

What matters is separating abandoned checkout from abandoned cart. Somebody who has already entered an email address and a delivery address is considerably closer to buying than somebody who only put something in the basket. Those two groups need different emails and different intervals.

And yes, this is the flow with the largest share of revenue. Which is why it also holds the most work.

Flow 3: Browse abandonment, but selective

Trigger: Product page viewed several times, no cart.

Content: A single email after about four hours with the product viewed and the two questions that come up most often about it.

Purpose: Pick up interest before it cools off.

The limit matters here, otherwise it feels creepy. We only trigger after the second visit, send exactly one email, and never more than one a week. A single page visit is curiosity, not a buying signal.

Flow 4: Post-purchase with a tracking condition

Trigger: Order completed, then the delivery event from the carrier.

Content: Shortly after the purchase, the confirmation with a realistic delivery time. After delivery, a how-to-use email. A few days later, the review request.

Purpose: Avoid support tickets and collect reviews without being annoying.

This flow is support work at the same time. Around 40 percent of all our tickets are "where is my order". A post-purchase flow triggered by real shipping events takes a large part of that away. The technical build behind it is in Automating WISMO requests.

The condition is the point. Without a delivery event you are guessing, and guessing produces exactly the emails that arrive at the wrong moment.

Flow 5: Replenishment at the point of use-up

Trigger: Purchase date plus the typical consumption period of the product.

Content: A reminder that argues from the product, not from a discount. With a toothbrush, that is the moment the head should be replaced.

Purpose: Repeat purchase at the right moment, without forcing a subscription.

At nano this is the flow with the best effort-to-return ratio. It is trivially built and has run unchanged for years. The only prerequisite is that you know the consumption period per product and have stored it in the shop.

If you have several products with different cycles, you need this as a product attribute. One fixed value for all products makes the flow inaccurate immediately.

Flow 6: Winback in two waves

Trigger: Last order lies well beyond the usual purchase interval.

Content: First wave without a discount, only with new products and a genuine question. Second wave two weeks later with an incentive, if there was no reaction.

Purpose: Bring inactive customers back and cleanly sort out the rest.

The second part is more important than the first. Anybody who does not react to either wave belongs out of the active list. A list full of dead contacts costs you money and also worsens your deliverability for everybody else.

Flow 7: Back in stock

Trigger: Customer signed up for a sold-out product, stock rises above zero again.

Content: A short email, immediately, with a direct link.

Purpose: Collect demand that would otherwise be lost.

The flow is unspectacular and still has one of the highest open rates of all. People want this email. They actively asked for it.

The condition behind it is a clean stock level. If your stock is out of sync between channels, you send emails for products that are already gone again. How to keep it accurate is in Multichannel inventory management.

What pays off poorly

So that you do not put your time into the wrong flow, here is the counter-list. We built all of these and switched them off again or cut them back sharply.

• Birthday emails. They rarely come across as personal. You have to ask for the date, many people give a false one, and the occasion has nothing to do with your product. For us the contribution was close to zero.
• Anniversary of the first order. The same problem in a different coat. Nobody celebrates the anniversary of a purchase.
• Five-stage cart chains. From the third email on, the return drops sharply and unsubscribes rise. Three is enough.
• Very fine-grained segments without a data basis. Twelve segments with 3,000 contacts means twelve groups that are all too small to prove anything. Three solid segments beat twelve theoretical ones.
• Prediction features on a small data basis. Purchase probabilities need volume. Below a few thousand orders that is tea-leaf reading with a nice interface.
• Pure AI copy without a brief. Fine as a draft, interchangeable as a finished product. Why we keep that strict is in Automating marketing in e-commerce and for product copy in Product descriptions with AI.

The limits

Three things no flow takes off your hands.

Flows do not replace an offer. If your product does not convince or your price does not fit, automation only amplifies the problem. It just sends it out more often.

Flows are only as good as the events underneath them. Five of the seven flows above depend on data from other systems: stock, shipping status, order history. If shop, carrier and email tool do not talk to each other cleanly, your best flow fires at the wrong moment. Here, n8n does that connecting work. The overview: Automating e-commerce processes.

More flows are not better. At some point a customer gets three emails in one day from three different flows. Set yourself a cap per recipient per week and check the overlaps before you build the eighth flow.

And a legal limit belongs here too: everything above assumes valid consent for marketing email. Transactional email such as order and shipping confirmations is a different matter from marketing. Mixing the two to get around consent will land you in trouble.

Conclusion

The seven flows above are no secret science. The difference lies in the conditions rather than the idea: delivery instead of fixed days, second visit instead of first, consumption period instead of a flat rate, a completed purchase ends the series.

Start with abandoned cart and post-purchase. Those two carry the largest share for us. Finish them before you start the third. One flow that runs beats five in draft mode.

If you want to know which flow gives you the most leverage first, and whether your data is even good enough for it: talk to us at Flowhouse.