Measurement & ROI: proving what your automation is worth
You’ve built flows — a welcome series, cart and browse recovery, a post-purchase sequence, lifecycle segments. Now the hard question: which of them actually made money, and how much? Marketing automation earns its keep only when you can point to the revenue it drove and the customers it kept. This lesson shows you how to measure that for a real store, using our running example, GlowKit, so you leave knowing exactly which flows to double down on and which to quietly retire.
We’ll keep the maths deliberately simple and hand-workable. You don’t need a data team to read your own automation — you need a few clear metrics, an honest attribution model, and the discipline to look at the same numbers every month.
What to measure (and what to ignore)
Automation platforms bury you in metrics. Most of them are vanity. Open rate, in particular, has become close to meaningless since mail providers began pre-fetching images and inflating opens — treat it as a rough deliverability signal, not a success measure. Click rate is more honest, but clicks don’t pay the bills. Revenue does.
Here is the short list that actually tells you whether automation is working for a store like GlowKit:
- Revenue per flow — total money attributed to each automated sequence over a period.
- Revenue per recipient (RPR) — flow revenue divided by the number of people who entered it. This is the great equalizer: it lets you compare a tiny high-intent flow against a big low-intent one fairly.
- Average order value (AOV) — are automated orders bigger or smaller than your baseline?
- Repeat purchase rate — the share of buyers who come back. For a replenishment brand this is the whole game.
- Customer lifetime value (LTV) — total gross profit a customer generates before they lapse.
- Flow-level conversion rate — of the people who entered a flow, how many bought?
Ignore raw send volume, list size as a goal in itself, and “engagement” scores that don’t connect to money. A list that grows while revenue per recipient falls is a list you’re degrading, not building. The test for any metric is blunt: if it moved, would you change a decision? If not, stop looking at it. Everything on the list above passes that test — each one, if it shifts, tells you to expand a flow, fix a flow, or protect a margin.
Attribution basics for D2C
Attribution is the rule you use to decide which touchpoint gets credit for a sale. There is no perfect answer — only a consistent one. The three models you’ll actually meet:
- Last-click — credit goes to the final touch before purchase. Simple, but it flatters bottom-funnel flows (like cart recovery) and starves top-funnel ones (like the welcome series that first warmed the buyer up).
- First-click — credit goes to the first touch. Flatters acquisition, ignores the flows that closed the deal.
- Time-window (click/open) attribution — the model most email and SMS platforms default to: a sale is credited to a flow if the customer clicked (or sometimes opened) a message from it within a set window, often 1–5 days for email and shorter for SMS.
For GlowKit, the practical approach is: pick one attribution window and hold it constant. If you use a 5-day click window for email, use it every month. Comparisons across months only mean something when the ruler doesn’t change. Be aware of double-counting: if a customer clicks both an SMS and an email before buying, two channels may each claim the full order. That’s fine for spotting which flows influence sales, but it means your summed “automation revenue” can exceed your true total. When you report an overall number, deduplicate to actual orders.
One more honesty check: separate automated flows from one-off campaigns (your Tuesday newsletter). This lesson is about the flows you built in earlier lessons — the automated email revenue flows that run without you touching them. Those are the compounding asset; campaigns are labour every time.
Revenue by flow, AOV & LTV
Let’s work the numbers. Everything below is illustrative — invented example figures chosen to teach the arithmetic, not GlowKit “data.” Plug in your own.
Revenue per recipient, flow by flow
Take a 30-day window. For each flow: attributed revenue ÷ recipients who entered = RPR.
| Flow (illustrative) | Recipients | Attributed revenue | Revenue per recipient |
|---|---|---|---|
| Welcome series | 4,000 | $12,800 | $3.20 |
| Cart recovery | 1,500 | $9,750 | $6.50 |
| Browse abandonment | 3,200 | $3,840 | $1.20 |
| Post-purchase / replenishment | 2,600 | $10,400 | $4.00 |
| Win-back (lapsed) | 1,800 | $2,160 | $1.20 |
Read that table and the story writes itself. Cart recovery earns $6.50 per recipient — it reaches people at the moment of highest intent, so a small audience prints money. Browse abandonment and win-back both sit at $1.20; they still contribute, but they’re where you’d experiment before you’d expand. RPR stops big flows from looking impressive just because they’re big. For context, welcome flows across ecommerce tend to be among the strongest revenue-per-message performers, which is why a $3.20 RPR here is unremarkable rather than magical.
AOV: are automated orders bigger?
Compute AOV per flow (flow revenue ÷ flow orders) and compare it to your store baseline. Suppose GlowKit’s overall AOV is $48 (illustrative). If cart-recovery orders average $65, the flow isn’t just rescuing sales — it’s rescuing your bigger carts, which raises its true value. If a flow’s AOV is well below baseline (common with heavy discount-led win-backs), factor in the margin you’re giving away before you celebrate the revenue.
Repeat rate and LTV: the number that decides the business
For a replenishment brand, acquisition is a cost; repeat purchase is the profit. Two quick formulas:
- Repeat purchase rate = customers with 2+ orders ÷ total customers, over a fixed window.
- Simple LTV = AOV × purchases per year × gross margin × average customer lifespan (years).
Illustrative worked LTV: $48 AOV × 3 orders/year × 0.60 gross margin × 2.5 years = $216 of gross-profit LTV per customer. Now watch what a flow does to it. If your post-purchase and replenishment automation lifts repeat orders from 3 to 3.6 per year, LTV rises to about $259 — a 20% gain that flows straight from automation you’ve already built. That’s the argument for measuring flows by their effect on retention, not just their first-order revenue. For grounding, published benchmarks put the typical D2C repeat purchase rate at roughly 25–30%, with skincare and consumables running higher thanks to natural replenishment — so a replenishment flow that nudges GlowKit up a few points is both realistic and valuable.
Reading the numbers
Now connect measurement back to money spent. Email as a channel is famous for a high return — industry estimates put average email marketing ROI at around $36–$42 for every $1 spent, and higher for well-run ecommerce programs — but a channel average is not your number. Compute your own:
- Flow ROI = attributed gross profit from the flow ÷ cost to run it (platform share + creative + any SMS/message fees).
- Automation’s revenue share = deduplicated automation revenue ÷ total store revenue. Track the trend, not the single reading.
When you read the table, ask three questions in order. First, what’s the RPR? That ranks your flows by efficiency. Second, what’s the volume? A high-RPR flow reaching only 200 people is a smaller prize than a mid-RPR flow reaching 4,000 — efficiency times reach is the real opportunity. Third, what’s it doing to LTV? A flow can look mediocre on first-order revenue while quietly driving second and third orders that never get credited to it.
Beware two traps. Discount leakage: a win-back that “recovers” $2,000 by handing out 25% off may be buying sales you’d have won anyway at full price — always read margin, not just revenue. Attribution inflation: if every channel claims every sale, your flows will collectively appear to drive more than 100% of revenue. Deduplicate before you brief anyone.
Iterating
Measurement is only useful if it changes what you do next. A simple monthly loop:
- Rank every flow by revenue per recipient.
- Expand the winners — more recipients, an extra well-timed message, a second variant. Cart recovery at $6.50 RPR deserves a third email before you build anything new.
- Fix or cut the laggards. A $1.20 flow gets one honest improvement attempt (better timing, a stronger subject line, tighter segmentation from your lifecycle segments) — then a decision.
- Test one thing at a time. Change the delay, or the offer, or the copy — not all three — so the number you move can be traced to the lever you pulled.
- Watch LTV quarterly. First-order revenue moves monthly; retention moves slowly. Judge lifecycle flows on the longer clock.
Give tests enough volume and time to reach significance before you call them. A flow touching 200 people a month can take a while to produce a trustworthy result — don’t kill it on a week of noise. A useful rule of thumb: wait until each variant has produced a few dozen orders, not just a few dozen clicks, before you trust the difference. And keep a change log — a one-line note of what you altered and when — so that three months from now you can explain why a flow’s numbers stepped up. Measurement without a memory just repeats the same experiments.
Your turn
Build your own revenue-by-flow view for one 30-day window:
- Set a single attribution window (e.g. 5-day click for email) and use it for every flow.
- For each automated flow, record recipients, attributed revenue, orders, and AOV.
- Calculate revenue per recipient and rank the flows.
- Calculate your store’s repeat purchase rate and a simple LTV, then estimate how much one retention flow moves it.
- Pick one winner to expand and one laggard to fix or cut — and write down the single change you’ll test next.
Free download: GlowKit Automation ROI / Revenue-by-Flow Tracker — a spreadsheet that calculates revenue per recipient for each flow and includes an LTV / repeat-rate calculator, so you can drop in your own numbers and rank your flows in minutes.
Once you can read which flows drive revenue and lifetime value, you’re ready to scale them without scaling your workload. Next up: using AI to expand and personalize what’s working. The Marketing Automation course hub keeps the full lesson map in one place.
zoho.social is an independent media platform and is not affiliated with, endorsed by, or associated with Zoho Corporation. All product names and brands are the property of their respective owners.
