# Direct Compost Solutions: how the search and AI model is built

Prepared by Team Empathy, 17 November 2026. Every figure below is the one in the strategy report. Pull it apart with your own AI assistant if you like; each step names its source.

## What the model measures

Non-branded Google organic search in Australia for what the store sells: compost bins and worm farms, compost worms, garden seeds, and soil minerals and conditioners. A non-branded search names the product and not Direct Compost, Compot, Eco-Eze or AB Seeds. Branded search, Google's shopping listings, paid search, social and email sit outside the model and are unchanged by it. All revenue figures are gross (before returns) unless marked net. Australian dollars throughout.

## Sources

- **Your Search Console** (sc-domain:directcompostsolutions.com), 17 July 2025 to 14 November 2026: clicks, impressions and positions by search, page, country and day.
- **Your Shopify**, 6 September 2026 to 17 November 2026 (the store's whole history): Google organic sessions and completed checkouts, orders, gross sales, returns and billing country.
- **Your Google Analytics**, the twelve months to 14 November 2026, mostly the old site: purchases by new and returning users.
- **Google Keyword Planner** volumes for Australia via DataForSEO, November 2026, with DataForSEO Labs' clickstream estimate as a cross-check.
- **DataForSEO** for authority, competitor traffic estimates and live Google results.
- **A live check** of every old-site address that earned clicks, 17 November 2026.

## Step 1: the market

- 260 Australian searches for what you sell, close variants counted once: **75,390 a month**.
- By territory: compost bins and worm farms 31,390, compost worms 4,980, garden seeds 35,310, soil minerals and conditioners 3,710.
- Long tail: x 1.35 for the smaller wordings below the cut = 101,775 a month, 1,221,300 a year.
- How-to questions (12,500 a month measured) are left out: they bring readers, and they feed the guides.

## Step 2: the winnable pool

- Searches a year x 0.20 = **244,260 winnable organic clicks**. The 0.20 is the share of searches that end as an organic click once ads, shopping grids, marketplaces and AI answers take theirs.

## Step 3: today, and before the move

- Search Console, Australia, 15 September to 14 November 2026 (61 days): 273 clicks, 40 of them on branded searches, leaving 233 non-branded. x 365 / 61 = **1,394 a year**, 0.57% of the pool.
- The year before the move, 10 September 2025 to 9 September 2026: 4,261 Australian clicks, 4,087 non-branded = **4,087 a year**, 1.67% of the pool.
- Searches Google does not name individually count as non-branded; branded searches are frequent enough to be named.
- The fall, with seasons cancelled out: 758 Australian clicks from 15 September to 14 November 2025 against 273 in the same weeks of 2026, down 64%. Every country: 3,331 against 444, down 87%.

## Step 4: the value of a click (Shopify)

- Conversion rate: 1.22%, Shopify's Google organic sessions that completed checkout (14 of 1,147). Order value: A$42.44, Google organic gross sales over orders (16 orders), before returns. Value of a click: A$0.52 gross.
- This is a small sample from a ten-week-old store. Shopify counted 1,147 Google search sessions in the time Search Console recorded 742 clicks, so the conversion rate is likely understated.
- New customers: 70% of purchases came from new users over the twelve months in Analytics (Shopify's own split is meaningless on a store where nearly every buyer is new to it). A new customer's first order is valued at the same A$42.44.
- First-year value of a new customer: first order x 1.15 = A$48.81. Three-year value: first order x 1.40 = A$59.42. Both multipliers are projected. Cross-check: returning-customer revenue in Analytics ran at 70% of new-customer revenue over the same twelve months.
- Net figures take out your Shopify return rate, 7.6% (returns over gross sales after discounts).

## Step 5: each 1% of the pool

| Item | Value |
|---|---|
| Clicks | 2,443 |
| Orders | 29.8 |
| Gross revenue | A$1,265 |
| Net revenue (7.6% returns) | A$1,169 |
| New customers | 20.8 |
| Three-year value of those new customers (projected) | A$1,237 |

## Step 6: the cases

- Conservative: the share you held in the year before the move, 1.67%. Recovery, measured.
- Mid and high: each comparable store's estimated Australian visits a year (DataForSEO, branded included) as a share of the same pool. The median of the eight is 17.2%. That estimate counts their branded and off-range searches, so it overstates what they win here: the mid case takes a quarter (4.3%) and the high case half (8.6%).

| Case | Share | Clicks / yr | Gross revenue / yr | New gross revenue on today / yr | Net revenue / yr |
|---|---|---|---|---|---|
| Today | 0.57% | 1,394 | A$722 | - | A$667 |
| Conservative | 1.67% | 4,087 | A$2,117 | +A$1,395 | A$1,957 |
| Mid | 4.3% | 10,503 | A$5,441 | +A$4,719 | A$5,029 |
| High | 8.6% | 21,006 | A$10,881 | +A$10,159 | A$10,057 |

| Case | New customers / yr | First-order revenue / yr (gross) | First-year value / yr (gross) | Three-year value / yr (gross, projected) |
|---|---|---|---|---|
| Today | 12 | A$504 | A$580 | A$706 |
| Conservative | 35 | A$1,478 | A$1,700 | A$2,069 |
| Mid | 90 | A$3,799 | A$4,368 | A$5,318 |
| High | 179 | A$7,597 | A$8,737 | A$10,636 |

Gross revenue in each case is the pool times the share times the conversion rate times the order value, multiplied on the unrounded inputs; new gross revenue is each case's rounded revenue less today's.

## Checks

- Today's A$722 of non-branded gross revenue a year sits well under what Shopify recorded from Google organic (A$679 gross in ten weeks, branded searches included, about A$3,400 a year), which is why the value of a click is called a first reading that probably runs low. The whole store took A$4,961 gross in those ten weeks, about A$25,000 a year.
- On DataForSEO's traffic model you get about 314 organic visits a month in Australia; the mid case is about 875 non-branded clicks a month, against Tumbleweed's estimated 3,474.
- Authority: yours is 31 out of 100 against a median of 21 for the eight stores, so the gains in the model come from your own pages.
