Run on evidence, not instinct.Introducing Mend — the first AI decision engine for attributionThe AI decision engine

Insights See what sold, who sold it, and why.

Revenue drawn with every post and email on the line beneath it. Partners ranked by what they sold, not by their followers. What buyers wrote, counted as people. Every number says how MISSIN worked it out.

Reads from

  • Shopify
  • Instagram
  • TikTok
  • YouTube

Coming soon

  • Meta
  • Google Ads

The Insights home for Halcyon Skin, a sample store. Press a number to swap the plot; hover one to see how it was worked out.

A number you can take to a meeting.

Every figure says how it was worked out, in a sentence. When someone asks how you know, the answer is already on the number.

How it is worked out

Orders whose strongest touch reached Evidence, and what they earned. Not the same as attributed: a post sitting on an order is not a cause named.

£2,857 explained£4,475 with no cause proven, which is a finding too

How partners are ranked, compared and valued

  1. The arrows are each partner’s move against the other ranking. After the bar, Each partner’s share of the attributed revenue minus their share of the followers. Above zero earns more than its reach; below zero is reach that does not convert.. Six sample partners, September.

    Rank partners by what they sold, not their followers.

    The biggest audience is rarely the biggest earner. Sort your partners by the revenue Mend attributed to them and the order changes, with each partner's move shown against the follower ranking.

    • PostPriya's skincare under £306 ordersEvidence
    • PostAisha's reactive-skin routine23 ordersEvidence
    • PostFreya's barrier routine12 ordersEvidence
    • PostPriya's autumn first look3 ordersPossible
    • PostLila's Glow Serum reel41 ordersEvidence
    • PostPriya's budget haul6 ordersCoincidence
    • PostFreya's serum first look2 ordersPossible
    • PostAisha's morning shelf3 ordersPossible
    • Evidence
    • Possible
    • Coincidence

    See which format works for your brand.

    Reels, stories, long videos and image posts, measured on your own buyers rather than on received wisdom. A format's bar is only as sure as the posts in it: proven orders solid, possible ones striped.

  2. Share of followersShare of attributed revenue

    Priya Shah holds 41% of the followers and 8% of the attributed revenue: 231k followers, 6 orders, £890.

    Catch the reach trap. Find the quiet earner.

    Put each partner's share of the followers beside their share of the revenue. Reach that never turns into orders shows as exactly that, and so does the small account that sells well above its size, likely the one worth more budget.

  3. Order #1042 · £50.40

    • PostLila's Glow Serum reelConverted64% · £32.26
    • SendAutumn launch emailAssisted22% · £11.09
    • PostFreya's barrier routineIntroduced them14% · £7.06

    Rolls up to

    • LCLila Carter64%£32.26
    • Autumn launch emailYour own send · 22%£11.09
    • FHFreya Holt14%£7.06

    Counted whole, the 3 touches would claim £151.20 of a £50.40 order. Split, they claim £50.40.

    Post Worth adds up to partner worth.

    Each post is credited its share of every order it is on, and a partner is worth the sum of their posts. Shares on one order add up to the order, so nothing is counted twice, and pounds appear only where there is evidence.

What buyers say, counted as people.

Free-text answers are read for what they were about, how the buyer felt, the problem they bought to fix and why they trusted the creator. Then they are said back as plain sentences, each person counted once, and No sentence under 5 answers, counts only under 25, and no line for a group of fewer than 3 people, which is a coincidence, not a finding..

What they're saying

48 answers read, each person counted once.

  • 21 customers (44%) found the Glow Serum amazing
  • 17 customers (35%) bought it for sensitivity or safety
  • 14 customers (29%) trusted the creator for being honest
  • 9 customers (19%) found delivery great
  • 9 customers (19%) trusted the creator for their expertise

What they bought it to fix

In their wordsSafety & sensitivity: reactive skin11Safety & sensitivity: redness6Results: dull skin5Appearance: uneven tone4

Why they trusted the creator

Eight universal concern headings and six reasons to trust a creator, the same in every industry. Why a recommendation landed, honest, expert or relatable, is the brief to hand the next partner.

Sample answers for Halcyon Skin, a sample store. The sentences are the product’s own templates.

Read from the tools you already run.

All integrations
  • ShopifyShopifyConnected
    • Orders and net revenue
    • Codes redeemed, the strongest evidence
    • Buyers' answers at checkout
  • StripeConnected
    • Orders and net revenue
    • Promotion codes redeemed
  • InstagramTikTokYouTubeInstagram, TikTok, YouTube
    • Posts on the revenue line
    • Followers, to rank against
  • KlaviyoConnected
    • Emails and texts on the revenue line
  • MetaMeta AdsComing soon
    • Ads and what they cost
  • Google AdsGoogle AdsComing soon
    • Ads and what they cost

Stop paying for followers.
Pay for what sold.