Polaris Launch — Campaign Brief

The read

Five things about the positioning before any of this sends. Blunt, because it matters more than the copy.

"PPC Management That Feels In-House" is a delivery-model claim, not a reason to reply. Nobody changes agency because the next one feels in-house. Keep it on the site, keep it out of every message.
Two credibility legs are missing at once. No published client names and no address. $[redacted]M managed & $[redacted]M generated are real and usable, but they're unattributed, so they can support a line in M3 and never lead. Get two named references or two logo permissions before this scales past a few hundred sends, or you convert inbox to call and stall on the call at the proof question.
The LinkedIn page says Founded 2025, 2-10 employees, one employee listed. Against a site claiming 18 years and $[redacted]M generated. That gap is real to anyone who checks. The proof is the founders' track record, so it has to be framed as theirs, and this campaign runs from Tim's personal profile, never the company page.
"We were tired of pretty reports that meant nothing" is the strongest sentence they own and it's buried under the tagline. This campaign is that sentence in practice: turn up with an idea, not a report. Retire "start with a free audit" from outbound, because it's the identical offer every PPC agency is sending this week.
Three published result sets, none attached to a vertical. The site shows 125%/44%/1,500%, 15%/21%/10% and 309%/15%/10% under "Real Results From Real Clients", and separately shows logos. The pairing isn't published, so we can quote a number against a vague descriptor but cannot say which industry it came from until Tim confirms it. That confirmation is a gate.

The shape changed on 12 Aug. The campaign no longer reports a fault it found in their account. It turns up with a creative idea built off one of their own products. That is a generative opener rather than a critical one, and it removes the entire false-positive problem in one move.

Why this offer

Three diagnostic directions were designed and argued against each other first: who is bidding on their brand, the Microsoft slots they don't hold, and where their live ads actually land. All three share one flaw. They open by telling a stranger something is wrong, which puts the reader on the defensive and puts the sender on the hook for being right.

Bringing an idea inverts that. Nothing has to be wrong for the message to land, nothing has to be verified twice before it can be sent, and the reader has no reason to argue. The three diagnostic angles stay in the drawer as reply material once a conversation is open.

The offer is a dangle, not a giveaway. We say the ideas exist and offer to walk through them. We do not attach them, link them or email them. The artefact is the reason for the meeting, not a substitute for it.

SPOTOne real product on their own site, named specifically.

SKETCHCreative ideas for it, anchored to a comparable published result.

WALKTwenty minutes. They keep whatever's useful.

Honest trade-off, stated plainly: this asks for a meeting rather than permission to send a file, so raw reply rate will sit below a pure give-it-away offer. It should convert far better per reply, because a walkthrough sells and an emailed PDF sits unopened. For a retained service at this price that is the right side of the trade.

The messages

Blank connection request. Sends from Tim's personal profile. No em dashes, no emojis. Nothing is claimed about their account, their spend or their numbers at any point.

M1 · the idea50 frame · ~62 filled

Hi {{first_name}}, I was looking at {{product}} on the {{company}} site and had a few ideas for ad creative, based on a campaign we ran for {{vague_descriptor}} that lifted {{result}}. I've sketched them out. Easier to show you than write them out here. Want me to walk you through them?

"Easier to show you than write them out here" is the line that does the work. It gives an honest reason for a call instead of a manufactured one, and it stops the reader asking why this can't just be emailed.

M2 · the second thing49 frame · ~61 filled

{{first_name}}, one more thing I noticed while I was in there. {{second_observation}}. That's usually the sort of thing that shows up in the creative before it shows up in the numbers. Still happy to run you through what I put together, twenty minutes and you can take whatever's useful.

"You can take whatever's useful" removes the obligation without weakening the ask. The second observation must be about their public storefront or ads, never a claim about performance.

M3 · the close36 words

{{first_name}}, eighteen years in paid search and over $250 million in revenue generated for the brands we work with. Say the word and I'll put twenty minutes in the diary to walk you through the creative.

The recon

This got much lighter when the offer changed. There is no two-date verification, no deliberate purge and no false-positive machinery, because no fault is being alleged. About six minutes a lead.

1 ARE THEY ACTUALLY ADVERTISING? Google Ads Transparency Center by domain. Zero live ads = DROP. We are pitching creative to someone who buys media. If they don't, there is no conversation. 2 PICK ONE REAL PRODUCT Their own site. A named product, from a category they clearly push: the hero product, a bestseller row, or the item their current ads point at. It must be nameable in three words and still be on sale on the day of sending. 3 MATCH TO A PUBLISHED RESULT Choose the closest of the three published result sets by business model, not by industry, since the industry pairing is not published. Record which set was used against which lead. 4 ONE SECOND OBSERVATION FOR M2 Something visible on their storefront or in their live ads: the offer in the ad copy, the landing page the ad points at, the format mix, a promotion still running. Public only. Never a claim about performance, spend or results. 5 LOG IT Product, result set used, second observation, date checked. One row per lead, so nothing gets reused or contradicted later.

The ideas have to genuinely exist before M1 goes out. M1 says "I've sketched them out" and a yes with nothing to show kills the campaign in one reply.

Targeting

TITLES Founder, Co-Founder, Owner, Managing Director, Ecommerce Director, Head of Ecommerce, Head of Digital, Head of Trading, Marketing Director, Head of Growth, Head of Performance Marketing, Head of Paid Media EXCLUDE Anyone agency-side: PPC Specialist, Google Ads Consultant, Freelance PPC, Fractional CMO, Agency Founder. Also SEO, content, social, dev, HR, finance, sales. SIZE 11-200 employees, roughly $2M-$[redacted]M revenue. The real qualifier is catalogue and spend, not headcount: 200+ live SKUs, a Merchant feed serving Shopping tiles, and 6+ live ads in the Transparency Center. Maps to their own form buckets $10-25k, $25-50k, $50-100k. Under $5k monthly spend: dropped. SECTORS Apparel · sporting goods and outdoor · supplements and wellness · cosmetics and personal care · home and furniture · pet supplies · consumer electronics accessories · food and beverage DTC · large-catalogue auto parts TENURE 6 to 36 months in role. GEOGRAPHY One market per batch. See the gate below.

Signals - what makes this timed rather than demographic

A creative first shown in the last 21 days. Someone is actively working the account this month, so creative ideas land on a live desk rather than a dormant one.
A burst of new creatives in one week. Usually a replatform, a feed rebuild or a new agency onboarding. All three mean creative is being rewritten right now.
New marketing or ecommerce lead in seat under 120 days. First-quarter vendor reviews are the most reliable buying window in this segment.
Open req for Ecommerce Manager, Performance Marketing or Paid Media. Paid is under review when the team is being rebuilt.
A new product line or collection launched in the last 60 days. New product, no proven creative yet, and the ideas have somewhere obvious to go.
Seasonal peak inside 8 weeks. Creative decisions get made on a dated calendar, so the timing argues for itself.

Before it sends

Tim confirms which result set belongs to which kind of businessThe three published result sets are not attached to a vertical anywhere public. Quoting one against a lookalike descriptor needs him to confirm the pairing is honest, in writing. Without it the message either goes vague or goes wrong, and wrong is unrecoverable with this buyer.

The creative ideas exist before M1Genuinely sketched, per lead, not a promise to think about it later. M1 says "I've sketched them out". A yes with nothing to show ends it.

One market per batchThe recon browser sits in one country at a time. A US session checking a UK brand's Shopping results sees the wrong ads. Confirm Polaris Launch's own operating market and run the checks from an IP in the prospect's market.

Nothing is attached, linked or emailedThe whole point is the walkthrough. If a reply asks for it by email, the answer is a time offer, not an attachment. Brief this before launch or the campaign quietly becomes a PDF drop.

No claim about their accountNothing about spend, budget, CTR, ROAS or settings. The product, the storefront and the live ads are public. Performance is not.

Proof stays inside the published set18+ years, $[redacted]M+ managed, $[redacted]M+ generated, and the three result sets against vague descriptors. No client names, no invented figures, no percentage attached to this prospect.