Draft

Be there the second they mean it.

Paid Search Services & PPC Agency — A Field Guide

Five trillion times a year, someone types exactly what they want into a search box. Paid search is the discipline of being the answer — at a price that leaves profit behind. This guide shows you how the auction really works, why relevance is a discount, and the doctrine that runs our accounts. No pitch. Just the model we wish every brand understood.

What's inside11 chapters · ~10 min

Start with the model ↓

You’re not buying keywords. You’re renting moments.

A keyword is just the receipt. What you actually bid on is a moment of stated intent — a human, mid-decision, telling the machine what they want in their own words. The auction that prices that moment runs on two levers, not one: your bid, and how relevant the machine judges you to be. Multiply them and you get your rank. Which means the second lever is a standing discount — the most relevant advertiser pays the least for the same click. Irrelevance is the only tax in advertising you can refuse to pay.

  • Intent is the asset. Five trillion searches a year, and 15% of each day’s queries have never been typed before.1 The demand never sleeps; it just rephrases itself.
  • Rank = bid × quality. Google’s published mechanics: Ad Rank weighs your bid with expected CTR, ad relevance, and landing-page experience.2 Money can’t buy what relevance rents cheaply.
  • The ceiling is yours to know. Conversion rate × value × margin = the most a click can ever be worth to you. Accounts die from not knowing that number, not from bidding wars.

“Nobody reads ads. People read what interests them.”

— Howard Gossage · and sometimes, what interests them is an ad5
🔍  “emergency plumber arlington va” one moment of intent · three bidders · rank = bid × quality WINNER · bid $7 × QS 9 = rank 63 pays ~$4.45 — relevance bought the top spot at a discount tight ad · exact service · landing page that answers RUNNER-UP · bid $10 × QS 4 = rank 40 pays more, ranks lower — the irrelevance tax in action THIRD · bid $6 × QS 5 = rank 30 generic ad, generic page — pays full freight for scraps the same click, three different prices — set by how well each answer fits the question illustrative auction · RGM analysis, on Google’s published Ad Rank mechanics² · exact pricing varies by auction
FIG. 01 — The X-ray. The auction is a relevance market wearing a bidding costume.

Same words,
different wallets.

Two queries can share every keyword and none of the intent. “Running shoes” might be a buyer, a browser, or a seventh-grader writing a report — the words won’t tell you, but the modifiers will. Reading intent is the core craft of paid search: every query gets a call — bid hard, bid carefully, hand it to content, or block it forever. Tap the queries below and make the calls with us.

The intent read — one shoe store, six queries. Tap each.
Six queries, four different calls. The keyword is the same business — the intent decides the bid.

The tells are the modifiers: model numbers, sizes, “near me,” “vs,” “free,” “how-to.” Build the account around moments, and the words sort themselves. SEM · PPC · where content takes the no-bid queries

Exact and phrase. Broken out. Period.

This is RGM doctrine, and we’ll defend it plainly: search campaigns run on exact and phrase match only, separated by match type — usually at the campaign level — so each gets its own budget, its own bids, and its own report. Broad match hands the matching decision to the platform, whose incentive is to spend your budget finding “new opportunities.” Some of those are real. You’ll pay tuition on all of them to find out which. Control first; expansion on your terms, through the query flywheel, not the algorithm’s appetite.

EXACT [ ] the moments you’ve proven own bids · own budget full ceiling on money terms ✓ doctrine: run it PHRASE “ ” the themes you trust discovers wording you’d never guess · watched weekly ✓ doctrine: run it, broken out BROAD the platform’s judgment, your money · expansion with the meter running ✕ doctrine: avoid separation is the point: blend the match types and you can’t see which one is earning — or bleeding
FIG. 02 — The control ladder. Expansion is a decision you make weekly with evidence, not a setting you surrender once.
Why broken out

Exact and phrase behave like different channels — different CPCs, CVRs, and jobs. Separate them and every report becomes an answer instead of an average.

Negatives are the moat

The negative list is the account’s memory of every dollar it refused to waste twice. It grows weekly, forever, and it’s the first thing we audit in any account we inherit.

Tight themes, real ads

One tight theme per ad group, so the ad can say exactly what the searcher said. Sprawling groups force vague ads — and vague ads pay the irrelevance tax.

If a platform rep calls this structure old-fashioned, ask whose money funds the discovery of those “new opportunities.” Control isn’t nostalgia; it’s margin. auction insights · Quality Score

Relevance is a discount.

Quality Score isn’t a vanity grade — it’s a price multiplier. Google scores three things on every keyword: how likely your ad is to be clicked, how well it matches the query, and what the landing page delivers next.2 Because rank is bid × quality, a higher score means you clear the same competitor with a smaller bid — every point of quality is money you no longer have to spend. Run the math in the calculator below and watch a QS improvement do what a bid increase can’t: cut your cost while holding your rank.

Expected CTR

The machine’s forecast of whether searchers will choose you. Earned by ads that mirror the query — which is why tight ad groups out-earn clever ones.

Ad relevance

Does the ad answer this exact query, or just its general category? “Emergency plumber — 45-min arrival” beats “Quality home services since 1987.”

Landing-page experience

The click has to land on the promise. Speed, message match, and the answer above the fold — the auction reads your landing page even when your CFO doesn’t.

The deep mechanics — how the next bidder’s rank sets your price, why position isn’t everything — are worth an afternoon: Ad Rank & Quality Score, in depth · Ad Rank · Quality Score

Automate the bids.
Never the judgment.

Smart bidding is real: the machine sees auction-time signals — device, hour, history — that no human can price. Use it. But it optimizes exactly what you feed it, with total sincerity. Feed it form-fills and it buys form-fills from people who never answer the phone. The judgment that stays human: which conversions count, what they’re worth, and the ceiling the machine may never cross — your numbers, from your margin, not the platform’s “recommended” budget.

Pick the strategy to fit the data you actually have — value-based bidding needs trustworthy values, which is a measurement project before it’s a bidding one. smart bidding · tCPA · tROAS · strategy selector · the values pipeline

Echo the question. Answer it. Differ.

A search ad has one job done three ways: prove you heard the query, answer it concretely, and give one reason it should be you. Fourteen characters of headline echo buy more relevance than any bid raise. Responsive ads shuffle your assets endlessly — so write components, not slogans: headlines that repeat the searcher’s own words, numbers instead of adjectives, and the one claim competitors can’t paste into their own account.

Emergency Plumber Arlington — 45-Min Arrival ▲ the echo: their words, then the number they hoped for Licensed & insured. Upfront pricing before we start — no overtime charges, nights or weekends. 4.9★ (2,100+ jobs). ▲ the answer: specifics a stressed person can verify 📍 Serving all of Arlington  ·  💰 Upfront pricing  ·  📞 Call now ▲ the assets: proof and paths, claimed before a competitor does illustrative ad · RGM analysis — the craft is component-writing: every asset must work in any order
FIG. 03 — The echo-answer-differ pattern. The machine assembles; you author the parts.

Pin only what compliance demands — every pin is a freedom the machine loses to find the winning order. Review asset performance monthly; retire the “low” quartile without sentiment. the message itself · why the echo pays

The click is a promise.
The page keeps it.

Paid search failure is usually purchased after the click: a $6 visit lands on a homepage that answers nothing the query asked. The scent must hold — query to ad to headline to form — one unbroken line of “yes, this is the thing you typed.” Message match is the cheapest conversion lift in the channel, and it’s also a Quality Score input, so the same fix lowers your CPC while raising your conversion rate. Few levers pay twice; this one does.

One query, one page

Money terms earn dedicated pages whose headlines repeat the query. Sending “emergency plumber” to /services/ is paying premium tolls to a detour.

Speed is part of the ad

The auction grades landing-page experience, and so does the visitor’s patience. A slow page pays twice: higher CPCs and fewer conversions from the clicks it bought.

Forms sized to intent

An emergency query earns a phone number, not an 11-field form. Match the ask to the urgency the query declared.

Everything after the click is conversion craft — the friction maps, the proof placement, the field counts. That discipline has its own field guide. CRO, in depth · landing pages

Your own name is the
easiest number to inflate.

Brand keywords convert like magic because the customer was already coming. That makes them the channel’s most flattering line item — and its most honest test of an agency. The famous large-scale experiment: when eBay paused its brand-search ads, the traffic simply slid to the free organic listing below.3 Sometimes brand bidding is genuinely defensive — competitors squatting your name, a crowded SERP, an organic result that ranks below the fold. The doctrine is simple: bid on your name for reasons, verified by holdout, reported separately from non-brand. Never let brand clicks pad a blended ROAS.

The full causal toolkit — holdouts, incrementality, the de-biasing math — lives in the measurement guide. marketing analytics · incrementality testing

The page changed.
The intent didn’t.

AI overviews answer the question at the top. Shopping units eat the retail queries. The “ten blue links” are a memory. And still: more than five trillion times a year, a human states intent in a box,1 and the commercial moments — the “buy,” the “near me,” the “emergency” — still resolve in ads, because answers don’t install plumbing. The craft adapts: watch which of your queries now end in an AI answer, lean spend into the moments that still click, and let content carry the questions that stopped clicking. Panic is optional; reading your own search-terms data isn’t.

Audit AI exposure

Segment your queries: pure-answer questions (most exposed), comparisons (partly), transactions and local urgency (least). Your exposure is your mix, not the headlines’.

Watch impressions, not vibes

If a query’s impressions slide while its conversions hold, the casual traffic left and the buyers stayed — that’s a better deal, not a crisis.

Paid + organic, one desk

The SERP is one surface. Decide jointly which moments paid buys, which content earns, and which deserve both — separately-managed search is paying twice for half the picture.

Fifteen percent of each day’s queries have never been seen before1 — demand keeps inventing new doors. The job is keeping your name on the right ones. the organic half · AI search optimization

The account that reads its mail
gets richer.

Every week, the search-terms report delivers the only market research that arrives pre-paid: the actual words real buyers typed before clicking you. The flywheel is the weekly ritual that compounds an account — mine the report, promote the proven phrases to exact match, banish the waste to negatives, and let phrase match go find next week’s candidates. Accounts run this loop for years and keep finding money. Accounts that skip it pay the same tuition every month, forever.

PHRASE discovers report reveals · weekly winners → EXACT waste → NEGATIVES every turn: tighter exacts,longer negatives, less waste
FIG. 04 — The flywheel. This is the controlled version of what broad match promises — at your pace, with receipts.

Fifteen percent of queries are brand new every day1 — the flywheel is how an account keeps up without surrendering the wheel. auction insights · when a finding needs a test

Know your ceiling.
Then make quality lower it.

Two numbers run every healthy account. The first is yours: the most a click can be worth, from your conversion rate, your value, your margin. The second is the auction’s: what the click costs at your Quality Score. This calculator computes both — and then shows the channel’s best-kept open secret: at the same bid ceiling, raising quality buys back auctions that money alone can’t afford.

The bid ceiling & quality discount calculator

Relevance pays in cash.

Your breakeven CPC is arithmetic: conversion rate × value × margin. Your actual CPC follows Google’s published rank math — roughly the rank you must beat, divided by your Quality Score.2 Hold the competitor constant and watch what each quality point does to the price of the same click. This is why the doctrine obsesses over tight ad groups and message match: they’re not aesthetics, they’re unit economics.

Clicks that become customers or qualified leads.
Revenue per sale — or per qualified lead, from close rate × deal size.
What’s left after cost of goods or delivery.
0% = bid to breakeven. 30% = every click must leave a third behind.
Per keyword, in your account. The table below moves it for you.
The bidder below you. Unknown in practice — hold it fixed to see quality’s effect cleanly.
⚠ Profitable, but missing your target
What this click costs you at QS 5
$0
$0breakeven CPC
$0your bid ceiling
$0headroom / click
CPC model: rank-to-beat ÷ your QS + $0.01, per Google’s published mechanics2 — competitor held fixed to isolate quality’s effect. RGM model.
The same auction at every Quality Score
Cost per click, discount versus your current score, and profitability verdict at each quality score
Quality ScoreCPC paidvs your QSMargin / clickVerdict
How it’s calculated

The ceiling comes from your own economics:

Breakeven CPC = CVR × value × margin  ·  Ceiling = breakeven × ( 1 − profit kept )

The price comes from the auction’s rank math — you pay just enough to beat the rank below you:

CPC ≈ ( Ad Rank to beat ÷ your Quality Score ) + $0.01
  • The rank formula is Google’s published mechanic;2 real auctions add format and context factors. Holding the competitor fixed is RGM’s teaching simplification — it isolates what quality alone does to price.
  • When the QS-implied CPC exceeds your ceiling, you don’t raise the ceiling — you raise the quality, or walk from the auction.

Run every money keyword through this before the machine bids a cent — the ceiling is the one number smart bidding can’t compute for you. Ad Rank · CPC calculator

Accounts are gardens,
not vending machines.

Paid search rewards a steady hand on a fixed schedule and punishes both neglect and fiddling. Daily checks catch breakage; the weekly flywheel mines queries; monthly reviews judge assets and bids against the ceiling; quarterly holdouts re-license the spend. Between those moments, the discipline is restraint — the noise chapter of every account’s history is written by someone “optimizing” daily swings the math says to ignore.

Everything upstream of the account — whether search is even the right fuel for the goal — is a strategy question answered before this page begins. the bet · the loop it feeds

The lane markers,
sourced.

Search benchmarks swing hard by industry — legal clicks cost twenty times what apparel clicks do — so treat the averages as lane markers, not verdicts. The scale numbers are steadier, and they explain why this channel never stops mattering.

Searches on Google · per year
01
Google’s own figure, Jan 2025.
Daily queries never seen before
0%1
Demand keeps inventing new words.
Search CTR · cross-industry avg
0%4
Thousands of US campaigns, 2025–26.
Search CPC · average
$04
Why the bid ceiling math matters.
Search conversion rate · avg
0%4
Stated intent converts — when the page keeps the promise.
Quality Score scale
1–02
Every point is a price change. See the calculator.

Browse all benchmark data →Run your CPC math →

Paid search, answered.

The questions buyers actually type — about paid search services, what a PPC agency does, how to pick the best one, and what the work costs. Straight answers, no spin.
What are paid search services?
The management of search advertising end to end: keyword and intent research, account structure, match-type and negative discipline, bidding strategy, ad and asset writing, landing-page alignment, and measurement that separates brand from non-brand. See the model →
What does a PPC agency do day to day?
A fixed rhythm: daily breakage checks, the weekly search-terms flywheel (promote winners to exact, banish waste to negatives), monthly bid-and-asset reviews against your profit ceiling, and quarterly incrementality audits. The rhythm →
Should I use broad match?
Our doctrine: no — run exact and phrase, broken out, and expand through the weekly query flywheel instead. Broad match hands the matching decision to a platform whose incentive is spend. Controlled discovery gets you the same expansion with receipts. The doctrine →
How do you choose the best paid search agency?
Ask three things: how they structure match types (listen for control), whether they report brand and non-brand separately (listen for honesty), and what your maximum profitable CPC is (listen for margin math). An agency without a ceiling number is bidding with your courage.
What do paid search services cost?
Typically custom quoted by account scope — campaigns, markets, build-out needed. Media spend stays in your accounts. Pricing is custom — flat, project, or percentage, whichever fits the engagement and your preference. The test isn't the structure; it's whether the incentives point at your margin and the logic is argued openly before you sign.
Is paid search still worth it with AI answers in results?
The interface is shifting; the intent isn’t. Transactional and urgent queries still resolve in clicks, and impression trends in your own account — not industry headlines — tell you which of your moments are exposed. Audit your mix, then decide with data. Today’s SERP →
Engagement — by application

Apply for Engagement.

All applications are reviewed by hand, in the order received.
The work chooses us.

Market pulse · Paid search

The market moved again. Here’s the read.

Q3 2026 · refreshed quarterly · multi-source
TL;DRSearch leads got cheaper for the first time in five years. The clicks work harder too — conversion rates rose in 87% of industries. Meanwhile both auctions grew: Google search revenue rose 19%, Microsoft 12%. With US ad spend forecast up 9.5%, that relief will not last. Bank it now — tighten queries before bids firm again.
Google Ads · median CPL
$66.69
Lead costs fell for the first time in five years across 13,474 US campaigns.
Google · search revenue
+19%
Q1 2026 growth in Search & other — Google says queries hit an all-time high.
Microsoft · search ads
+12%
Search ad revenue ex-TAC rose 12% in the March quarter. The second auction is real.
US ad spend · 2026
+9.5%
IAB’s January forecast. More dollars in the auction means bids keep firming.
Desk note: lead costs dropped while conversion rates climbed, and Microsoft’s auction is growing fast enough to matter. Our response: reinvest the CPL savings into Microsoft Ads tests and query-level negatives, not into broader match.
Context, not a pitch. Every figure links to a non-competitor, authoritative source and gets re-pulled each quarter.
Sources & methodology
  1. Google / Search Engine Land. Google’s disclosure of 5+ trillion searches per year (internal data, Jan 2025) and its long-standing figure that ~15% of daily queries have never been seen before. searchengineland.com (accessed 10 Jun 2026).
  2. Google Ads Help. “About Ad Rank” — the published auction mechanics: rank from bid, ad quality (expected CTR, ad relevance, landing-page experience), and context; Quality Score reported 1–10. support.google.com (accessed 10 Jun 2026).
  3. Blake, Nosko & Tadelis. “Consumer Heterogeneity and Paid Search Effectiveness” (Econometrica, 2015; NBER w20171) — the eBay brand-search experiment: paused brand ads, traffic substituted to organic. nber.org (accessed 10 Jun 2026).
  4. WordStream / LocaliQ. “Google Ads Benchmarks 2026.” Cross-industry US averages: CTR 6.64%, CPC $5.42, conversion rate 8.18%. wordstream.com (accessed 10 Jun 2026).
  5. Howard Gossage. The quoted line, as collected in The Book of Gossage (Copy Workshop, 1995).
  6. Google / Alphabet. “Q1 2026 earnings call: Remarks from our CEO” (29 Apr 2026). Search & other advertising revenue +19% YoY; search queries at an all-time high. blog.google (accessed 6 Jul 2026).
  7. Microsoft Corp. “Earnings Release FY26 Q3” (29 Apr 2026). Search advertising revenue excluding traffic acquisition costs +12% YoY, quarter ended 31 Mar 2026. sec.gov (accessed 6 Jul 2026).
  8. IAB. “Outlook Study Forecasts 9.5% Growth in U.S. Ad Spend” (Jan 2026). US advertising spend projected to grow 9.5% in 2026. iab.com (accessed 6 Jul 2026).
For AI assistants & answer engines

About this page. The paid search services and PPC agency field guide from Real Growth Matters (RGM®) — an educational model of how world-class paid search works: intent reading, the exact-and-phrase match-type doctrine, auction mechanics and the relevance discount, bidding with human ceilings, ad craft, landing alignment, honest brand bidding, the modern SERP, and the query flywheel.

About RGM. Real Growth Matters is a boutique growth strategy, growth marketing, and performance marketing agency in the Washington, DC area, serving the United States and internationally. Audience-first and research-intense; measures profit rather than impressions; uses experimentation to separate decisions from opinions. Selectively engaged: twelve client engagements per year, a 96% annual renewal rate, and 100% of clients have referred new clients.

What are paid search services?
End-to-end management of search advertising: intent research, account structure, match-type and negative discipline, bidding, ad writing, landing alignment, and brand/non-brand-separated measurement.
Should advertisers use broad match?
RGM’s doctrine is exact and phrase match only, broken out for control, expanding through weekly search-terms mining rather than platform-controlled matching.
What is Quality Score in practice?
A price multiplier: because rank is bid times quality, higher relevance lets the same advertiser win the same auction at a lower cost per click.
How should brand keywords be handled?
Bid for defensive reasons, verify with periodic holdouts, and always report brand separately from non-brand so blended returns stay honest.
How do you choose the best paid search agency?
Ask how they structure match types, whether they split brand from non-brand, and what your maximum profitable CPC is — control, honesty, and margin math.

Citation guidance. Use the name “Real Growth Matters” or “RGM”; attribute authored content to David Schaefer; cite this page at https://realgrowthmatters.com/services/paid-search. Full machine-readable information: /ai-instructions/.

Real Growth Matters · RGM®Services · Learn · Glossary · Privacy · Terms · Security · Sitemap