Where things stand
We went through your store, both of your ad accounts and your analytics. This is the position as your own systems record it, covering 3 June to 31 August 2026. Nothing here is an opinion you cannot check.
| Search visibility | WEAK | Absent for the searches a guitarist with the problem actually types. | |
| Visibility in AI answers | ABSENT | No independent source exists anywhere for an assistant to cite. | |
| Competitive position | EXPOSED | The most expensive of four. The other three do not advertise at all. | |
| Creative system | ABSENT | 44 distinct images. No test was set up between any two of them. | |
| Brand and visual direction | ABSENT | No repeated palette, type or logo position across those 44. | |
| Media strategy | SCATTERED | 7 campaigns, 3 objectives, 15 ad sets, 5 optimisation goals. | |
| Budget concentration | THIN | About $338 a month, divided across all of the above. | |
| Mobile experience | WEAK | Price sits below the fold. The buy button is 1.6 screens down. | |
| Store conversion | WEAK | 2,410 sessions produced 88 add to carts and 8 orders. | |
| Discounting | UNCONTROLLED | A 25% launch sale and a signup offer that moved between 10% and 15%. | |
| Measurement | BROKEN | $435 of revenue never recorded. No server side sale event exists. | |
| Retention | ABSENT | No post purchase sequence. 19 orders, and no reason to return. | |
| Unit economics | INVERTED | About $296 to win an order worth $94.12. |
Five blocks is what selling this product online asks for, not what is theoretically possible. Every line is evidenced below, and every figure comes from your own systems between 3 June and 31 August 2026.
Most people never find you at all
"Guitar string protection case" is not a phrase people search for. The category does not exist in their heads yet, the way "capo" or "guitar case" does. For the searches a guitarist with rusting strings actually types, you do not appear anywhere. On the narrower searches where you do appear, you sit behind general instrument case retailers, an Etsy listing and a direct competitor. Separately, 33 of your 180 organic visits land on an old Portuguese path that returns an error today.
| What someone searches | Where you rank |
|---|---|
| "how to protect guitar strings" | nowhere |
| "keep guitar strings from rusting" | nowhere |
| "guitar string protector" | 5th of 9 |
| "guitar string case" | 7th of 8 |
If they ask an assistant instead, it recommends someone else
We ran three real queries against a live AI assistant on 2 September. The pattern is the same each time: there is no independent source anywhere for it to cite. On your own brand name it found only your pages, and said so unprompted. Nothing about this is a ranking problem you can fix with keywords. Nobody but you has written about your product.
| Query | What came back |
|---|---|
| "String ProTech guitar reviews" | Your own pages only. It noted it could not find independent reviews. |
| "why do guitar strings rust so fast" | Eight sources, all forums and articles. None of them you. |
| "best guitar string protector case" | You appeared 6th of 7, below a competitor it called the first of its kind. |
Three other people solve this problem, and none of them buys ads
Your competitors reach buyers two ways you currently do not: shelf space in shops people already walk into, and content that answers the question before the product is mentioned. You are the most expensive of the four, you are the only one advertising, and you are the only one selling exclusively from your own site. That combination is why paid traffic has to carry everything.
| Price | Sold through | Advertising | |
|---|---|---|---|
| You | $59.99 | Your own site only | Yes |
| The String Sling | $34.95 | Own site, Guitar Center, Target, Music & Arts | None found |
| The Guitar Necktie | $45.00 | Its own site only | None found |
| G-Suit | ~$40 | Site no longer resolves | Not checked |
So you buy the attention instead
Every image that has run as an ad, together. There is no repeated frame, no recurring typeface, no consistent colour and no fixed position for your logo. Six of them ran as three separate ads each. One audience of 1,008 people saw the same ad an average of 18 times, 18,087 impressions for $459, which your account average of 1.89 hides completely.
3
3
3
3
3
3
2
2
2
2
And the buying is asked to do five different things
Eight of your fifteen ad sets were never asked to produce a sale. They were asked for post engagement, a video view, or a profile visit. On roughly $338 a month, the system buying your traffic was given five definitions of success at the same time.
| What the ad set was told to optimise for | Ad sets |
|---|---|
| Offsite conversions | 7 |
| Post engagement | 3 |
| Messaging purchase conversion | 3 |
| Thruplay, meaning video views | 1 |
| Profile visit | 1 |
They land, and seven in ten land on a phone
On a 390 pixel screen the first thing they see is a logo and a photograph. Your price is 68 pixels below the fold. Your buy button is another 416 pixels below that. The first place you reassure anyone they can get their money back is 96 percent of the way down a page that runs 10.8 screens long. On a desktop the price and the button sit in the right hand column, visible the moment the page loads. Same store, different experience, and mobile is 69 percent of your traffic and converts at 2.5 percent against desktop's 8.0.
Your best pages are the ones that are not selling anything
The pages that explain hold people comfortably. The pages that ask for money do not. The case page loses the average visitor in fifteen seconds.
| Page | Sessions | Bounce | Avg. time |
|---|---|---|---|
| About | 55 | 10.9% | 1m 55s |
| Blog | 25 | 12.0% | 7m 44s |
| Bundle product page | 1,356 | 57.4% | 1m 19s |
| Case only product page | 104 | 64.4% | 0m 15s |
They meet two different discounts on the way
Your announcement bar offers 25 percent off a limited time launch sale that is ending soon. It has been ending soon for at least three days. Your signup popup offered 15 percent on one visit and 10 percent on another. Someone deciding whether $79.99 is fair is being told three different things about what the real price is.
They buy, and a quarter of it is never recorded
Your analytics recorded $1,353.38 against the $1,788.30 you actually took. Your ad platform never receives a server side copy of a sale, so it has been optimising on partial information this whole time. And nothing happens afterwards: there is no post purchase sequence, and no way for the 2,402 people who did not buy to hear from you again.
| Step | Count | Of all sessions |
|---|---|---|
| Sessions | 2,410 | |
| Added to cart | 88 | 3.7% |
| Reached checkout | 22 | 0.9% |
| Purchased | 8 | 0.3% |
What that order actually cost you
Spread every dollar of advertising across every paying order you have ever taken, whether an ad caused it or not, and each order cost about $296 to win. It is worth $94.12. That is the most generous reading available: it credits advertising with your organic, direct and email sales too. Your ad platform's own stricter count puts it at $436. Even at a theoretical 100 percent margin, with the case costing you nothing to make, break even would be $94.12, and you are more than three times above it. This is why the sample being small does not rescue the number. It is not close.
| Gross margin | Break even cost per order | Current $296 is |
|---|---|---|
| 50% | $47.06 | 6.3x over |
| 70% | $65.88 | 4.5x over |
| 90% | $84.71 | 3.5x over |
| 100%, product free to make | $94.12 | 3.1x over |
- A product people keep. 19 paying orders and no refund problem
- Pages that hold attention: about, blog, contact
- An audience that engages when it arrives on its own
- A year of advertising history in two accounts
- A live store on a maintained theme
- Any brand guideline or visual system
- A creative testing plan
- A content calendar
- Server side purchase tracking
- A post purchase email flow
- One campaign objective, chosen and held
- A mobile first product page
- A discounting policy
- Search content answering what buyers ask
The right hand column is not a criticism of effort. It is the reason $5,623.18 has not told you what you needed it to. Eight purchases in ninety days cannot prove your product wrong, and they cannot prove it right. But the cost per order is not a sample size problem: at more than three times the ceiling, it is not close, and no amount of extra data at this setup would close that gap. What the numbers cannot tell you is whether a properly built version of this performs differently. What they can tell you is that the version that ran was never arranged to find out.
- Make the numbers trustworthy. Server side sale tracking, the missing quarter of revenue, and the dead paths. Until this is done every judgement after it is a guess, including whether the next three worked. Days, not weeks.
- Fix the page that receives the traffic, mobile first. Price and buy button above the fold, reassurance near the decision instead of near the footer. Every extra visitor sent to the current page costs more than it needs to.
- Answer the question before you sell the product. Your competitors do not buy ads. They rank for what a guitarist with rusting strings actually types, and you are absent from every one of those searches. This is also the only route to an independent source existing for an AI assistant to cite. It is the slowest item, so it is the one to start soonest.
- Then run one advertising channel properly. One objective, creative tested against itself, budget concentrated rather than spread, pointed at a page that is ready. Done in this order, the cost per order has somewhere to fall to. Done first, it produces another $5,000 of ambiguity.
Your unit cost on the case, which is what turns revenue into margin. Where two of your campaigns actually send a click, which is five minutes inside your Ads Manager. And whether Google spend predates May 2026, since there is April traffic with no cost recorded against it, so $821.77 is a floor rather than a total.