There is a reason to be careful about turning ads on at Safe Haven Metal before the plumbing is right. Two things are currently broken in ways that would waste every dollar spent. Orders are not all reaching Google Analytics, so campaigns would be judged on incomplete numbers. And Merchant Center has no live price connection, so products get pulled for price mismatch the moment prices move. Both are fixable, both are quick, and both have to come first.
This covers our work only. Advertising spend is paid directly to Google and is not included here. After three months of tuning there is a genuine decision point: the numbers either justify continuing and increasing, or they do not, and we will have spent a modest amount finding that out rather than a large one.
Where the effort goes
Stages 1 and 2 are prerequisites and can run at the same time. Stage 3 is the launch itself. The ongoing month to month work is the largest single piece, because getting a campaign live is the easy part and making it pay is not.
Tracking & MeasurementStage 1
$2,330–3,415
Merchant Center & Price FeedStage 2
$2,145–3,415
Campaign Build & LaunchStage 3
$1,725–2,760
Three Months of TuningOngoing
$3,840–5,760
Stage detail
Stages 1 and 2 have no dependency on each other and can be done in parallel. Nothing should launch until both are finished and verified.
Stage 1
Tracking & Measurement
$2,330–3,415
Before any spend · Runs alongside Stage 2
- Orders are going missing, and we can prove it. Checking your recent transactions against Google Analytics, at least one real order did not appear. If that is happening at any regular rate, every campaign report will understate what advertising actually produced, and the natural response is to switch off something that was working.
- Find out why, then fix it. We export the transaction numbers from WooCommerce and from Analytics and compare them directly. That tells us whether this is a tracking code problem, an order status problem, or simply customers running ad blockers, which is common in this audience and is worth knowing about because it changes how the numbers should be read.
- Sort out Tag Manager. There are several containers in the account and it is not obvious which one is live or who owns them. One of them carries a pending invitation from the agency you used years ago. We consolidate to a single container that we know is correct.
- Confirm the purchase conversion is real. A campaign optimises toward whatever you tell it to value. If the purchase event is misfiring or reporting the wrong amount, Google will spend your budget chasing the wrong thing and do it efficiently.
- A dashboard you can actually read. One page pulling Analytics and Google Ads together, so performance is visible without logging into three systems and interpreting them.
Range breakdown
Conversion tracking audit & reconciliation$740–1,110
Tag Manager cleanup & consolidation$555–740
GA4 e-commerce event repair$555–925
Reporting dashboard$480–640
Access to Google Ads, Merchant Center and Tag Manager is already granted.
What moves the range
What the audit turns up. If the missing order is an isolated ad blocker case, this is confirmation work and lands at the low end. If the purchase event is genuinely misfiring, it has to be rebuilt and re-verified against live orders, which is the high end.
Stage 2
Merchant Center & Price Feed
$2,145–3,415
Before any spend · Runs alongside Stage 1
- This is the one that stops the ads working. Your product details reach Merchant Center through the feed plugin, and that part is fine. What is missing is anything that keeps the price current through the day. When the price Google has does not match the price on your site, Google disapproves the product and the ad stops running. On bullion, where the price moves constantly, that is not an occasional problem. It is the normal state.
- A direct price connection to Google. We add a light connection that pushes current prices to Merchant Center on a schedule, separately from the main product feed. It carries prices only, so it is small and puts very little load on your server. This is what lets the ads stay live.
- Clean up the feed itself. Attributes, titles and product categories get reviewed so that products are eligible in the first place and match what people search for.
- Work through the disapprovals. Google's policies treat gold bullion as close to currency, which means approvals in this category are inconsistent even when everything is correct. Some of this is resubmission and appeal rather than a technical fix, and it is worth knowing that up front rather than being surprised by it.
Range breakdown
Merchant API price sync build$1,110–1,850
Feed mapping & attribute cleanup$555–925
Disapproval review & policy resolution$480–640
Assumes the existing feed plugin stays in place for product details. The price connection supplements it rather than replacing it.
What moves the range
How many products fight back. A clean catalog with a handful of disapprovals is the low end. A catalog where a large share are flagged under the currency policy takes real back and forth with Google, and that is the high end.
Stage 3
Campaign Build & Launch
$1,725–2,760
After Stages 1 and 2 are verified
- Performance Max, built around margin rather than volume. Bullion margins are thin, which means a campaign judged on revenue alone can look excellent and lose money. Products get grouped by what they actually contribute, and the campaign is pointed at the groups that can carry the cost of a click.
- A return target set before launch, not after. We agree the minimum return on ad spend that makes this worth doing, and the campaign is configured to hold to it. That number is the whole test.
- Start small and contained. A controlled budget on a narrow product set, so the first month buys information rather than inventory movement. Widening comes after the numbers say it should.
- On suggested spend, somewhere around $2,000 to $3,000 a month is enough to get readable data in this category without risking much. Less than that and the campaign never gathers enough to learn from.
- Gold IRA is deliberately not in this stage. It is the highest value thing you sell and also among the most expensive clicks on Google, at figures like $67 each. It deserves its own test with its own landing page and its own budget cap, and it should come after the e-commerce side is proven to measure correctly. Doing both at once would make it impossible to tell which one worked.
Range breakdown
Product segmentation & ROAS targets$555–740
PMax build, assets & feed setup$800–1,280
Conversion value & bidding configuration$370–740
Advertising spend is billed by Google directly and is not included in these figures.
What moves the range
How many product groups get built and how much creative is needed. One tightly focused campaign is the low end. Several groups with separate targets and assets, which gives more to learn from but takes longer to build, is the high end.
Ongoing
Three Months of Tuning
$3,840–5,760
$1,280 to $1,920 per month · Then re-evaluate
- Launching a campaign is not the work. The work is the weeks afterward: seeing which products earn their clicks, cutting the ones that do not, adjusting targets as the data thickens, and watching for disapprovals as prices move.
- Monthly reporting against the agreed return target, in plain numbers. What was spent, what came back, and what changed as a result.
- Feed maintenance is part of it. Products get disapproved, prices drift, and new products arrive. Left alone, a working campaign degrades on its own.
- Then a real decision. At the end of three months the numbers either justify continuing and spending more, justify a different channel, or say that paid search does not work at these margins. All three are useful answers, and the point of doing it this way is to reach one of them cheaply.
- Other channels stay on the table. Email to past customers costs nothing per click and tends to outperform paid advertising for repeat purchases. Reddit communities such as the silver and gold groups are where this audience genuinely spends time. Neither is in this plan, and both are worth revisiting at the three month mark.
Range breakdown
Monthly management, reporting & feed upkeep$1,280–1,920
Three months$3,840–5,760
Month to month rather than a fixed term. Stops whenever you decide it should.
What moves the range
How much is running. A single campaign on a narrow product set sits at the low end. Multiple campaigns, a wider catalog and active disapproval management sit at the high end.
Total investment
Planning range · MRA work
Stage 1, tracking and measurement Before any spend$2,330–3,415
Stage 2, Merchant Center and price feed Before any spend$2,145–3,415
Stage 3, campaign build and launch After 1 and 2 are verified$1,725–2,760
Setup subtotal One time$6,200–9,590
Three months of tuning $1,280 to $1,920 per month$3,840–5,760
Setup plus three months$10,040–15,350
Plus advertising spend, paid to Google directly$2,000–3,000 / month suggested
What we need, and what is not included
To get started
- A starting monthly budget for Google spend.
Not included
- Advertising spend. Billed by Google directly to you.
- Gold IRA campaigns. Worth testing separately once the e-commerce side is measuring correctly.
- Reddit and community work. Genuinely promising for this audience, and better judged after three months.
- Email and repeat customer marketing. Likely the cheapest revenue available to you, and worth its own conversation.
- Search engine optimisation and content. A longer play than this plan covers.
- Brand campaigns. Not yet worth running, since there is little search volume on the name to defend.