Paid ads with a plan,
a gate, and proof.
Campaigns reviews the business and its ad accounts and writes the plan, audits measurement before any budget moves, checks every campaign before a penny, watches every day and proposes changes for you to apply, and proves the result in sentences. On Google Ads, Meta Ads, Microsoft Advertising, LinkedIn Ads, TikTok Ads and Pinterest Ads.
Example proposals, to show the flow. Nothing here moves until a person clicks.
What is Campaigns?
Campaigns is paid media management software from xenzee for agencies, in-house marketers and founders. It covers the whole job, not the dashboard: it reviews the business and its ad accounts and writes the plan, audits measurement before any budget moves, builds one campaign into a version per network and checks each before a penny, watches every day and proposes changes with evidence for a person to apply, and proves the result in sentences a business owner can follow. It buys on Google Ads, Meta Ads, Microsoft Advertising, LinkedIn Ads, TikTok Ads and Pinterest Ads.
- A review that writes the plan, with evidence
- Measurement audited before anything moves
- Every campaign checked before a penny
- Proposals, never autopilot
- The result in sentences, every figure sourced
Who is Campaigns for?
One account or a book of clients: the same review, the same gate, the same daily proposals, the same report.
Agencies
A project per client, the desk across every connected account sorted by need, client review links for approvals, and a report that answers the fee conversation with what was done. Adding a client is adding a project.
In-house marketers
Targets derived from the margin and shown as working, a tracking audit that can say fix first, and a Monday digest and a monthly report your finance team can read without a glossary.
Founders and freelancers
The review says what to run and why, the gate catches what a second pair of eyes would, and the proposals say what to change each day. You decide; nothing spends without you.
What does Campaigns do?
Six groups, the same six a project is organized in. Each line is a tab or a page you can open today.
Review
Start from the business and the account as they are, not from a media plan template.
Review
Start from the business and the account as they are, not from a media plan template.
- The paid review
- Fourteen phases, from the commercial model to the roadmap: what a lead or a sale may cost from the margin, where demand is, which network does which job, whether measurement can be trusted, how each account is built, what is being wasted, and what to do first. Every finding names its evidence.
- The plan
- A monthly line, a learning floor and a target per network, each held for four weeks before the next change, with the reason written down. The watch and the builder read the plan; nobody moves money against it by accident.
- Read the whole account
- Every campaign, ad group, ad, keyword and conversion action already in the account is read in, not only what was built here, so the review and the watch see what the client sees.
- Re-review on a cadence
- The review runs again each quarter and after any large change, and writes to one client record shared with the suite's SEO and social apps.
Measure
If the numbers cannot be trusted, nothing else is allowed to move.
Measure
If the numbers cannot be trusted, nothing else is allowed to move.
- Tracking audit, fix first
- Conversion actions, tags, auto-tagging, the website's tags, the tagging contract and the CRM link are read and checked. A verdict of "fix first" blocks the building and budget phases until the signal is right.
- One tagging contract
- Every network gets the same source, medium and campaign tags with its own click macros, and the click ids ride into the suite CRM with each lead.
- Outcomes back to the network
- When the CRM marks a lead qualified or won, the outcome is reported back to the network that sent the click, once, with the value the CRM recorded, so the network learns from real results, not form fills.
- The Money Map
- Spend, clicks, conversions, visits, leads, qualified, won and revenue in one table per network, every figure labelled with where it came from: the network's own count, modelled, analytics, CRM, or not connected.
Build
One window from the plan to every network, paused until a person presses go.
Build
One window from the plan to every network, paused until a person presses go.
- Brief from the client record
- Offer, proof, voice, competitors, brand terms and locations come from the shared client record; you change what you disagree with.
- Build once, tailor per network
- One base becomes a version per network shaped by that network's rules and the review's findings: consolidated search with negatives, broad social with enough creative to learn, a lead form on LinkedIn, vertical video on TikTok.
- Before a penny
- Proofreading, policy, tracking, landing page, budget floor and ceiling, negatives, duplicates and brand safety are checked on every version. A corrected version is offered, never applied for you.
- Client approval, then paused
- A member or a client approves from a review link; anything the gate would stop cannot be sent for approval. Every network publishes paused; you activate.
- Edit what exists
- Budgets, statuses and negatives on campaigns the account already had, through the plan's floor and ceiling, each change with a reason on record.
Watch
Read every day, propose with evidence, apply with one click, undo for a week.
Watch
Read every day, propose with evidence, apply with one click, undo for a week.
- Proposals, not autopilot
- Pacing against the plan, cost drift, spend with no result, disapprovals, account health, learning resets, search terms that waste, creative fatigue. Each anomaly becomes a proposal with the evidence, the expected effect as a range, the risk, and one action you apply. Nothing moves on its own.
- Hourly where it matters
- Pacing, disapprovals and the account's standing with the network are read every hour and written on the account.
- Armed rules with a boundary
- The narrow rules a person arms, with a person-set limit the plan enforces, and every firing logged as a proposal.
- The weekly loop and skills
- One change a week at most, in a fixed order, with the reason. Saved routines built from steps the app already performs, whose output always lands as a draft or a note.
Prove
What we spent, what it brought, what we changed and why, what is next, what we did for you.
Prove
What we spent, what it brought, what we changed and why, what is next, what we did for you.
- The report in sentences
- Five sections written from the numbers on the page, every figure traceable to its source, scheduled by email with a PDF. A narrator may reword the sentences; it may not add or drop a number.
- Tests, labelled as tests
- Brand on-and-off and holdout tests proposed when spend justifies them, run on a written schedule through the change log, read as a range from complete weeks, and marked TEST on the report.
- The change log
- Every change made from here, by a person or an armed rule, with its reason, alongside what the network changed on its own.
- One record across the suite
- Spend and results flow to the suite's CRM, project and reporting apps as live reads, never samples.
Operate
The routine runs itself, up to the point a person is needed.
Operate
The routine runs itself, up to the point a person is needed.
- The desk
- Every connected account across your projects, one row each, sorted by how much it needs you, with one next action that links to the page that does it.
- The suspension desk
- When a network restricts an account or disapproves an ad: the exact notice in the network's words, the likely cause, the appeal path, the rule never to open a second account, and an appeal pack with the evidence gathered.
- Roles and review links
- Owner, admin, editor and viewer per project; invited people join the suite with one link and never pay; client review links for approvals.
- Caps you can see
- Projects, ad accounts and a fair-use spend line per plan, shown with usage on the plan page. Spend above the line is never stopped.
What does Campaigns check before a penny is spent?
Every version of a campaign is read as a reviewer would read it, and the landing page is opened, before it can be sent for approval and before it publishes, paused. A version the gate stops is held with the reason on it. The corrected text is offered, never applied for you.
- Copy reads cleanSpelling and grammar in every headline, description and primary text, per network. The corrected line is offered, never applied for you.
- Policy, named per networkClaims that need proof, personal attributes, phone numbers and emoji where a network forbids them, competitors named by accident, regulated verticals and their disclosures, the words behind most suspensions.
- Tracking is in placeThe tagging contract on the destination, auto-tagging on the account, a conversion action the plan can optimise to. If the audit says fix first, the build waits.
- The landing page answersThe page is fetched: it answers, it repeats the words of the headlines, it is not the homepage, and it carries a privacy policy and a way to make contact.
- Budget inside floor and ceilingEach version's daily budget against the network minimum, the plan's learning floor and the monthly ceiling across every version together.
- Negatives and brand exclusionsA generic search campaign carries the brand and intent-drift negatives; a brand campaign does not exclude its own name; Performance Max is reminded of its brand exclusions.
- No duplicate campaignA campaign with the same name already live in the account stops the version.
- Brand safety notedPlacements and inventory settings the network offers are named so a person can decide.
A version the gate held
Example verdict, to show the flow.
How does Campaigns use AI?
AI does the reading, the first draft and the wording. You decide what ships. Nothing AI produces here spends, publishes or changes an account on its own.
- Reads the business, words the reviewThe website, the client record and the accounts become findings with evidence and a synthesis in plain words. The numbers in it are computed, never written by the model.
- Drafts copy per networkHeadlines, descriptions and primary text in the brand's voice from the client record, shaped to each network's limits, then proofread. Every piece carries its disclosure flag.
- Proofreads before the gateEvery line is read for spelling, grammar and the wrong word. The correction is offered on a click; the model's own slips are corrected and named before a draft reaches you.
- Words the report, keeps every numberThe report's sentences are computed from the numbers; a narrator may reword them into paragraphs but may not add, drop or change a figure, a label or a date. When it is unavailable, the computed sentences ship and the page says so.
The report, worded from the numbers
Between 1 Sep and 30 Sep the account spent 3,880 across two networks. Google took 2,700 of it for 1,912 clicks and brought 61 leads, 19 of them qualified in the CRM. Meta took 1,180 for 2,340 clicks and 24 leads, 5 qualified. The plan's monthly line is 4,000; spend ran at 97% of it.
Two changes were made from here: fourteen search terms were excluded on 12 Sep because they had spent with no result, and the Meta budget was raised to the floor on 19 Sep because the network was under its line.
Example paragraphs, to show the shape. Every figure on a real report is labelled with its source.
A pile of separate tools, or Campaigns
Most paid media is run from each network's own tool, a spreadsheet and a dashboard export. Here is what changes when the whole job lives in one place.
| What you are comparing | Separate tools | Campaigns |
|---|---|---|
| Where the targets come from | A benchmark someone read once | The client's own margin and close rate, shown as working |
| Trusting the numbers | Whatever the network dashboard says | A tracking audit that can say "fix first", and a trust label on every figure |
| Building a campaign | Each network's own tool, one at a time | One window, a version per network, checked before a penny, published paused |
| Catching a mistake | The client spots the typo in the live ad | Copy, policy, tracking, landing page and budget checked before approval |
| Day to day | Alerts, or an autopilot you cannot explain | Proposals with evidence and one click, undo for a week; nothing moves alone |
| A suspended account | A support form and a new account | The notice, the likely cause, the appeal path, the pack, and the rule: never a second account |
| Proving it worked | A dashboard export | What we spent, what it brought, what we changed and why, what is next, what we did |
| Adding a client | Another login in every tool | Another project |
How does Campaigns work?
A flow one person can run alone and a team can follow, for one account or fifty.
Connect and read
Enter the website, connect the ad accounts over each network's own sign-in, and let Campaigns read everything already in them.
Review and plan
The paid review derives the targets from the margin, audits measurement, reads each account and writes the plan: a line, a floor and a target per network. Change what you disagree with.
Build, check, approve
One brief becomes a version per network. The gate checks each before a penny; a client approves from a link; every network publishes paused and you activate.
Watch and prove
Every day the watch proposes with evidence and you apply with one click. Every month the report says what was spent, what it brought, what changed and why, what is next, and what was done.
The networks Campaigns buys on
A network is on this page only once Campaigns can read the account and publish to it. Networks still being verified are not listed.
Google Ads
Search, Performance Max, Display, Demand Gen, YouTube
Meta Ads
Facebook, Instagram, Threads
Microsoft Advertising
Bing search and the Microsoft Audience Network
LinkedIn Ads
B2B campaigns with lead forms tied to CRM outcomes
TikTok Ads
Vertical video, or nothing
Pinterest Ads
Search and feed placements
Is Campaigns safe to trust with client accounts?
Behind the pages there is an engine whose whole job is keeping the accounts inside their networks' rules, the money inside the plan, and every change explained.
- Connected the official wayEach network is connected over its own sign-in. Campaigns never sees or stores your platform passwords. Tokens are encrypted at rest and refreshed before they expire, and you can disconnect any account in one click.
- Separated at the databaseEvery organization's data is separated by row-level policies in the database itself, not only in the application, and roles per project decide who can connect accounts, launch spend and see results.
- Money moves only inside the planEvery live change goes through the network minimum, the plan's floor and the monthly ceiling, needs a reason, and lands in the change log with an undo for seven days. Every campaign publishes paused.
- Restrictions handled, never worked aroundThe account's standing with each network is read every hour. A restriction opens one alert with the notice, the likely cause and the appeal path, and the rule never to open a second account is on the page.
Account health, read hourly
Example accounts, to show the panel.
Questions people ask about Campaigns
What is Campaigns?
Campaigns is paid media management software from xenzee for agencies, in-house marketers and founders. It reviews the business and its ad accounts and writes the plan, audits measurement before any budget moves, builds one campaign into a version per network and checks each before a penny, watches every day and proposes changes with evidence for a person to apply, and proves the result in sentences a business owner can follow. It buys on Google Ads, Meta Ads, Microsoft Advertising, LinkedIn Ads, TikTok Ads and Pinterest Ads.
Which ad networks does Campaigns support?
Google Ads, Meta Ads, Microsoft Advertising, LinkedIn Ads, TikTok Ads and Pinterest Ads. A network is listed here only once Campaigns can read the account and publish to it. Networks still being verified are not on this page.
Does Campaigns change my budgets or bids on its own?
No. The daily watch reads the accounts and turns each anomaly into a proposal with the evidence, the expected effect, the risk and one action. A person applies it, and can undo it for seven days. The only exceptions are the narrow rules a person arms with a boundary the plan enforces, and every firing is logged as a proposal. Nothing is published live: every campaign lands paused and a person activates it.
What does Campaigns check before a penny is spent?
Spelling and grammar in every line, policy risks named per network, the tagging contract and a conversion action, the landing page (fetched and read for message match, a privacy policy and contact details), the daily budget against the network minimum and the plan's floor and ceiling, negatives and brand exclusions on search, duplicates already live, and brand safety settings. A corrected version is offered and applied only when you click it. A version the gate would stop cannot be sent for client approval.
How does Campaigns use AI?
AI reads the business and the accounts for the review and words its synthesis, drafts ad copy per network, proofreads every line, and rewords the client report from the computed sentences without adding or dropping a number. A person decides what ships. When AI is unavailable the built-in checks and sentences still run and the page says which it was. Provider and model choices are the platform's to make and change, not something your account depends on.
What happens when measurement cannot be trusted?
The tracking audit gives a verdict. "Fix first" blocks the building and budget phases of the review until conversion tracking, tagging and the CRM link are right, and the report labels every figure with its source until then. Campaigns does not optimise toward a number it cannot stand behind.
What happens if a network suspends an account?
The hourly read notices the change and the project's owners get one alert with four things: the exact notice in the network's words, the likely cause, the appeal path for that network, and the rule never to open a second account. An appeal pack gathers the evidence a reviewer asks for: the business facts, what the website shows, the ads in question with the pre-check on their words, every change made with its reason, and the pre-launch checks on record. You send the appeal through the network's own path.
Is Campaigns good for agencies with many clients?
Yes. Every client is a project with its own accounts, plan, review, builds, watch and report, and owner, admin, editor and viewer roles. The desk lists every connected account across projects by how much it needs you, client review links carry approvals, and the report answers the fee conversation with what was done, in sentences.
Can a founder or a freelancer use it alone?
Yes. One person gets the same review, the same gate, the same daily proposals and the same report, with no team required. The proposals say what to do and why; you decide.
How does Campaigns keep ad accounts safe?
Each network is connected the official way, over its own sign-in; Campaigns never sees or stores your platform passwords. Access tokens are encrypted at rest and refreshed before they expire. Every organization's data is separated at the database level. Every live change goes through the plan's floor and ceiling with a reason on record, and every campaign publishes paused.
Does Campaigns connect to the rest of the xenzee suite?
Yes. One xenzee account signs you into every app. A project here is the same client record the suite's SEO and social apps use, so the brand, competitors and offers are entered once. Leads and their outcomes come from the suite CRM, and spend and results are read live by the suite's project and reporting apps.
Can I look around before I commit?
Yes. Start an account and explore. Connecting an ad account and running the review needs an active plan; the plan page shows every cap and the usage against it, and spend above the fair-use line is never stopped.
Run paid media the way you would explain it to the client.
A plan from the business, a gate before a penny, proposals with evidence, and the result in sentences. Setup takes an afternoon.