Five percent of the sending does forty percent of the work
When a month comes in soft, the instinct is to add a Tuesday send. Klaviyo's 2026 benchmark across more than 183,000 brands says that is close to the worst move available, because the leverage is not in volume. It is in six triggered messages most companies have never finished building.
The same list, two completely different economics
A campaign is a message you decide to send. A flow is a message a customer's own behavior triggers. They use the same list and the same software, and they perform nothing alike.
Klaviyo's 2026 benchmark report, drawn from over 183,000 brands and an analysis spanning hundreds of billions of emails, puts the split starkly. Look at the two bars below and notice they are the same width.
Share of sends against share of revenue
Same programs, same lists. The top bar is effort. The bottom bar is return.
Klaviyo 2026 Omnichannel Benchmark Report, 183,000+ brands. Revenue per recipient was $1.94 on flows against $0.11 on campaigns, roughly an eighteenfold gap. Placed order rate ran 2.11% on flows against 0.16% on campaigns, and click rate 5.58% against 1.69%.
Omnisend's separate dataset tells the same story with different numbers, putting automated emails at roughly 2% of sends and around 30% of revenue. Two vendors, two methodologies, one conclusion.
The mechanism is not mysterious. A flow arrives because somebody just did something: joined a list, viewed a product, started a checkout, placed an order, or went quiet for ninety days. The message is relevant because the timing is. A campaign arrives on Tuesday because it is Tuesday.
Three flows do most of the work
Within the automated category the distribution is also lopsided. Welcome, abandoned cart and browse abandonment together account for roughly 87% of all automated email orders. That is the whole reason this is a tractable problem: you are not building forty things, you are building three properly and then four more.
There is an obvious counterpoint and it deserves stating. Flows fire at people already showing intent, so some of that revenue would have arrived anyway. Anyone quoting flow revenue as pure incremental gain is overstating it, and the honest way to find out is a randomized holdout, which we run whenever the volume supports it.
One more finding that surprises people: nearly 48% of flow-driven revenue comes from new buyers, against 16% for campaigns. Automation is usually filed under retention. It is doing a substantial share of first-purchase acquisition too.
Which flows are live, and what the missing ones are worth
Switch on the flows you genuinely have running today, not the ones somebody set up once and never checked. Then look at what the gaps cost you.
Flow coverage audit
A rough model using published per-recipient benchmarks. Nothing is stored or sent.
Flows you have live today
The model applies published per-recipient benchmarks to the trigger volumes your numbers imply. It is deliberately conservative and it is still a model, not a forecast for your account. Average order value, category, offer strength and list quality all move the result, and vendor benchmarks are drawn from ecommerce brands, which flatters some businesses and not others.
What it is good for is ranking. If two flows are missing, the arithmetic tells you which to build first, and that ordering is far more reliable than the absolute numbers.
Sending more is now actively dangerous
In February 2024 Google and Yahoo began enforcing requirements on bulk senders, defined as roughly 5,000 messages a day to personal accounts. Microsoft followed with its own enforcement in May 2025. These are not best practices any more. They are conditions of delivery.
That changes the calculation on volume. Every additional campaign to a disengaged list raises your complaint rate and lowers your engagement signals, and both feed the reputation that decides whether your next message reaches anyone at all. The Tuesday send does not just underperform. It can quietly damage the flows that were working.
SPF, DKIM and DMARC
All three authentication records are required for bulk senders. Missing DMARC is now the most common reason a previously fine sender suddenly starts landing in spam.
One-click unsubscribe
A working list-unsubscribe header, honored within two days. Making it hard to leave now costs you more than the subscriber you were trying to keep.
Spam complaints under 0.3%
Google's guidance is to stay below 0.1% and never reach 0.3%. That is roughly three complaints per thousand delivered, which a stale list will exceed faster than most people expect.
Monitor by provider, not in aggregate
Gmail, Yahoo and Microsoft behave differently and fail separately. A healthy overall delivered rate can hide a specific provider quietly filtering you out entirely.
A sunset flow, which nobody enjoys
Automatically suppressing subscribers who have not engaged in months. It shrinks your list and improves almost every metric you care about, including revenue.
Warm a new sending domain slowly
Volume ramped over weeks, starting with your most engaged subscribers. The welcome flow is the ideal warm-up vehicle because it reaches people at their highest-engagement moment.
Six flows, in the order we build them
Not because six is a magic number, but because these six cover the moments where a customer has already told you what they want. Everything beyond them is refinement.
| Flow | Fires when | Why it earns its place |
|---|---|---|
| Welcome series | Someone joins your list | Typically the highest revenue-per-recipient flow in an account. The first 48 hours is the most engaged window a subscriber will ever give you, and the engagement you earn there sets the sender reputation governing every later send. |
| Abandoned checkout | Checkout started, not completed | The highest-intent trigger available. Somebody entered payment details and stopped. Two messages minimum, and the first should arrive within an hour. |
| Abandoned cart | Item added, checkout never reached | A separate and earlier signal than checkout abandonment, and frequently merged with it by mistake. Abandoned cart emails see notably higher open rates than the same brand's campaigns. |
| Browse abandonment | Product viewed, nothing added | The weakest signal of the three and still part of the trio that accounts for roughly 87% of automated orders. Requires a light touch, since the intent is real but early. |
| Post-purchase | An order is placed | Thank you, set expectations, request a review, and suggest the natural next purchase. Drives second-order velocity, which is where customer lifetime value is actually made. |
| Winback | No purchase in 60, 90 or 120 days | Cheaper than acquiring a replacement, and it doubles as list hygiene: whoever ignores the winback sequence is a candidate for suppression before they start costing you deliverability. |
What comes after the six
Replenishment for anything consumable, on a cadence matched to how fast people actually run out. Back-in-stock, which is the highest performer almost nobody runs. Price drop for saved items. Review requests timed to arrival rather than to purchase. VIP recognition for your best customers, which mostly means acknowledging them rather than discounting to them.
For businesses that are not ecommerce, the same logic maps cleanly. A quote that was sent and never accepted is an abandoned checkout. A service due for renewal is a replenishment flow. A lead that went quiet after a call is a winback. The triggers change name and the principle does not.
The gap between having a tool and using one
Roughly 85% of business-to-business marketers acknowledge they are not using their automation tools to anything like their full potential, and only a small minority of revenue operations professionals say they trust their own data accuracy. Those two facts explain most underperforming programs better than any creative critique would.
Nobody fails at automation because they picked the wrong platform. They fail because triggers were never finished, suppression logic was never written, and the data feeding the segments is wrong in ways nobody has checked. That is unglamorous work and it is the work.
Sources: Klaviyo 2026 Omnichannel Benchmark Report covering more than 183,000 brands, including flow and campaign share of sends and revenue, revenue per recipient, placed order rate and click rate; Omnisend 2025 and 2026 ecommerce marketing reports; Google and Yahoo bulk sender requirements effective February 2024 and Microsoft enforcement from May 2025. Benchmark figures are vendor-published, drawn largely from ecommerce senders, and should be read as directional rather than as predictions for an individual account.
Open rate stopped meaning anything, and reporting has not caught up
Apple Mail Privacy Protection pre-loads images regardless of whether a person opened the message, which registers as an open. Published averages now sit in the high thirties and low forties partly because of that inflation rather than because anyone reads more email than they used to.
The practical consequence is that open rate is no longer comparable across providers, across years, or against a benchmark. It is still the number most reports lead with, and it is still the number most agencies quote when a program is underperforming, because it is the easiest one to make look good.
Four numbers that still mean something
Click rate. Unaffected by privacy pre-loading, because a click requires a person. The best available proxy for whether the message worked. Benchmark data puts flows around 5.6% against 1.7% for campaigns.
Revenue per recipient. Normalizes across list sizes, so a flow sent to two hundred people can be compared fairly against a campaign sent to twenty thousand. This is the number that exposes the eighteenfold gap between flows and campaigns.
Placed order rate. What share of recipients actually bought. Roughly 2.1% on flows against 0.16% on campaigns in the benchmark data, with top performers at double those figures.
Spam complaint rate, by provider. Not a marketing metric so much as a survival one. Watch it separately for each major inbox provider, because an aggregate figure can look healthy while one provider is filtering you out entirely.
And one number nobody reports
Incremental revenue, measured against a holdout. Every platform will tell you what a flow was attributed. None will tell you what it caused, because that requires deliberately withholding the message from a random slice of people who would have received it.
It costs a little revenue to run and it is the only way to know whether an automation is earning its place or simply standing where the money was already going to walk past. We recommend it on any flow big enough to produce a readable result, and we accept that it occasionally makes our own work look less impressive than the dashboard did.
What a marketing automation engagement covers
Roughly in this order, and the first two frequently produce more than everything after them combined.
Deliverability first
Authentication records, complaint rates by provider, bounce handling and a sunset policy. There is no point building flows that land in spam, and this is usually cheaper than anyone expects.
The six core flows
Built properly, with real timing and real suppression rules rather than a template switched on and forgotten. Content that reflects your actual business, not the platform's sample copy.
Segmentation that means something
Engagement tiers, purchase behavior and lifecycle stage. Most accounts have dozens of segments and use three, which is a symptom of segments built before anyone decided what they were for.
Text where it fits
Short, timely and consent-based. Text works for order updates, appointment reminders and genuine urgency, and burns goodwill fast when used as another broadcast channel.
Data plumbing
Getting the events your flows depend on into the platform reliably. When a trigger misfires it is almost always this, and it is invisible until somebody checks. Custom work sits under web app development.
Holdout testing
Withholding a random slice so you can tell what the automation caused rather than what it merely observed. The only honest way to value a flow, and rarely offered.
Campaign strategy, list growth and the creative side of sending live on our email marketing page. If the constraint is that too few people ever enter the list in the first place, the answer is lead generation rather than automation, and we will say so. Where the flows are feeding an online store, the margin analysis behind them sits on ecommerce marketing.
Everything is judged on incremental revenue rather than attributed revenue, which is the same discipline behind our work as a conversion rate optimization agency. Attributed numbers flatter automation more than any other channel, because flows fire at people who were already leaning toward buying.
When automation pays, and when it does not
Worth doing if
- You have a list and fewer than four flows genuinely running
- Customers buy more than once, so lifecycle timing exists to exploit
- Your open rates have been sliding and nobody has checked authentication
- You send campaigns weekly and cannot say what any flow produced
- You have never suppressed a disengaged subscriber
- Somebody set up flows two years ago and nobody has opened them since
- You sell anything that runs out, expires or renews
Not yet if
- Your list is a few hundred people, where manual beats automated
- Customers buy once and never again, with no referral motion
- You want more sending volume rather than better triggers
- Nobody can supply the purchase and behavior data flows depend on
- You want a platform migration, which is a project rather than a strategy
- You expect automation to fix a product or pricing problem
- You are not willing to remove anyone from the list, ever
Why us
Search Converts has been building marketing systems since 2012, and we run automation on our own portfolio of web properties as well as for clients. That matters because we have made the specific mistakes this page warns about, including sending too often to a list we had not pruned and watching the deliverability consequences arrive a month later.
We also sell every marketing channel, which is the reason to trust the recommendation. An email agency concludes you need email. We regularly conclude that your flows are fine and your list is not growing, which is an acquisition problem, or that the offer is the constraint and no sequence fixes it.
The founder additionally runs a home-services business outside the agency, which is a useful test of whether this thinking survives outside ecommerce. Most published automation benchmarks come from online stores. Applying them to a business where the transaction happens on a phone call requires translation rather than copying, and we do that translation rather than pretending the numbers transfer.
There are no invented case studies here and no borrowed logos. Every figure is published research with the source named and its limitations stated, including the point that vendor benchmarks are drawn largely from ecommerce and that attributed flow revenue overstates incremental gain. The first useful thing we do is look at your account, count the live flows and check your complaint rate, and that costs nothing.
Marketing automation questions, answered with the benchmarks
It builds and maintains the messages that fire automatically when a customer does something, rather than the ones you schedule. In practice that means triggered flows across email and text, the segmentation and suppression rules behind them, the data plumbing that makes triggers fire reliably, and the deliverability work that decides whether any of it arrives. The measure of a good one is whether they can tell you what each flow produced incrementally, not just what got attributed to it.
By efficiency, dramatically. Klaviyo's 2026 benchmark across more than 183,000 brands found flows made up 5.3% of sends and around 41% of email revenue, with revenue per recipient of $1.94 against $0.11 for campaigns, roughly an eighteenfold gap. Placed order rate ran 2.11% against 0.16%. Omnisend's separate data tells the same story at roughly 2% of sends and 30% of revenue. Campaigns still produce the majority of total revenue because of sheer volume, so this is an argument about where the next hour of work goes, not about abandoning campaigns.
Welcome, abandoned checkout and abandoned cart, then browse abandonment, post-purchase and winback. Welcome, abandoned cart and browse abandonment together account for roughly 87% of all automated email orders, so the concentration at the top is extreme. Welcome is usually the highest revenue-per-recipient flow in an account and also does double duty, since engagement in the first 48 hours after signup helps set the sender reputation that governs every message you send afterward.
No, and this is the most important caveat on the page. Flows fire at people who have just shown intent, so some of that revenue would have arrived without the email. A cart abandoner may well have come back anyway. Attributed flow revenue therefore overstates the true gain, sometimes considerably. The honest way to measure it is a randomized holdout, withholding the flow from a random slice of eligible people and comparing outcomes. We run that wherever volume supports it, and we would treat any agency quoting attributed flow revenue as pure profit with caution.
Since February 2024, bulk senders, defined as roughly 5,000 messages a day to personal accounts, must authenticate with SPF, DKIM and DMARC, offer one-click unsubscribe honored promptly, and keep user-reported spam complaints below 0.3%. Google's guidance is to stay under 0.1% and never reach 0.3%. Microsoft began its own DMARC enforcement in May 2025. These are enforced conditions rather than recommendations, and a previously healthy sender can start landing in spam without any change to their content.
Usually not, and this is where instinct and evidence diverge sharply. When results dip, the reflex is to add a send. But additional campaigns to a disengaged list raise complaint rates and depress engagement signals, both of which feed the reputation deciding whether future messages arrive at all. That means the extra Tuesday send can degrade the flows that were working. If you want more revenue from email, the higher-return move is almost always finishing your trigger coverage and suppressing people who stopped paying attention.
Because an unengaged subscriber costs you more than they are worth. They never open, which drags your engagement metrics down, and they are disproportionately likely to mark you as spam when they finally notice you. Inbox providers read both signals. A sunset flow that automatically suppresses people after months of no engagement typically improves open rates, click rates, deliverability and revenue at the same time. It feels like going backwards and it is the single most reliable fix in a struggling program.
Yes, with translation. Most published benchmarks come from online stores, so the specific numbers will not transfer to a service business, and anyone telling you otherwise has not thought about it. The logic maps cleanly though. A quote sent and not accepted is an abandoned checkout. A service due for renewal is a replenishment flow. A lead who went quiet after a call is a winback. The triggers are different and the principle, that behavior-triggered messages outperform scheduled ones, holds.
Almost certainly the one you already have. Platform choice matters far less than trigger coverage, and migrations consume months that would be better spent building the flows you are missing. Roughly 85% of business-to-business marketers acknowledge they are not using their current tools to full potential, which suggests the constraint is rarely the software. We will recommend a change when there is a genuine capability gap, and that is uncommon.
Faster than most channels, because the audience already exists. A welcome flow starts producing as soon as the next person signs up, and abandonment flows begin firing immediately. Expect early signal within two to four weeks and a fair read within a quarter, since some flows depend on a lapse window of 60 to 120 days before they even trigger. Deliverability repairs take longer, because sender reputation recovers gradually rather than the moment you fix the record.
For specific jobs, yes. Order and delivery updates, appointment reminders, and genuinely time-limited things people asked to hear about. Text has extraordinary open rates and a correspondingly short patience threshold, so used as another broadcast channel it burns goodwill quickly and generates opt-outs you cannot recover. The consent requirements are also stricter than email and worth getting right the first time. Our general rule is that text should carry information the recipient would have wanted even if you were not selling anything.
Yes. Search Converts is based in Denver and works with businesses across the United States and Canada. Automation work runs entirely on platform access, shared reporting and calls. Call +1 720-712-8615 and we will count your live flows, check your complaint rate and tell you which single gap is worth the most, whether or not you hire us.
Tell us which flows are live and we will find the gap
Your platform, roughly how many subscribers you have, and which automations are genuinely running. We will tell you which missing flow is worth the most and whether deliverability is quietly costing you more than any of them.
Or call and we will go through the account:
+1 720-712-8615Your information is never sold or shared. We respond within one business day.
