Creative drafting
Generates headlines, primary text, and angle variations in your brand voice from a single brief.
How-to — — by Mahmoud Zalt
Learn how to run paid ads without a marketer using Sistava AI Employees: setup, creative, targeting, bidding, and a weekly routine that prevents budget burn.
Yes, with three constraints in place. First, cap the daily budget so a bad week costs less than a dinner out, not a month of runway. Second, run one platform at a time so attribution stays simple. Third, lean on an AI marketing employee to write the variants, build the audiences, and read the data every morning, because the part that ruins non-marketers is not strategy, it is forgetting to look at the account for four days while a broken ad eats the budget. Founders who do these three things land their first profitable campaign in four to six weeks, even with zero ads experience. The category does work. The shape most non-marketers run it in does not.
The smallest safe setup is one platform, one campaign, one audience, two creatives, and one daily cap. The instinct most non-marketers have is the opposite: spin up Google plus Meta plus LinkedIn, multiple audiences, four creatives each, and let the platforms optimize. That works for agencies with five-figure budgets and a media buyer watching the account hourly. It does not work for a solo founder with a few hundred euros and a busy week. The smallest setup costs you nothing in optionality because you can add the second platform once the first one is honest. What it buys you is a clear signal: when something goes wrong, you know exactly which creative, audience, and day broke. With a sprawling launch, the same broken ad costs three times more and tells you nothing.
Most of the daily work yes, the judgment calls no. A Sistava AI marketing employee can draft headline variations, write primary text in three angles, build audience definitions from your buyer profile, propose bidding strategies, and rewrite a losing ad based on the data the account is showing. What it should not do alone is decide whether to scale a campaign or change landing-page pricing, because those are business decisions, not media decisions. The clean split: the AI employee handles volume work like drafting fifteen creative variants in an hour and reading the dashboard every morning, you decide which winner gets the next budget bump. That split is what makes paid ads feasible for a solo founder, not a heroic act of learning everything yourself.
Generates headlines, primary text, and angle variations in your brand voice from a single brief.
Translates your buyer profile into platform-ready audience definitions and exclusions.
Recommends manual vs auto bidding, target CPA, and budget caps based on conversion data.
Reads the account every morning and posts a plain-English summary of what changed and why.
Spots underperformers, drafts a fresh angle, and queues the next creative for your approval.
The split between you and the AI employee changes the math of paid ads for a solo founder. Without help, every hour inside an ads dashboard is an hour not spent on product or sales. With an AI marketing employee handling the volume work, you trade thirty minutes of decision time per week for a function that used to need a part-time freelancer at roughly fifty times the cost. The first month still costs a learning budget on the platform, but the ongoing operational cost lands inside a flat monthly subscription rather than an hourly rate that scales with chaos.
Hiring the AI marketing employee is the easy part. The harder part is feeding it enough context that the creative does not read like a stock ad: who your buyer is, what they search for the night before they convert, which objections come up in calls, what offers have failed before. Spend the first session loading that context, not optimizing a campaign. Campaigns get better when the context is good, not when the bidding gets clever. That hour on day one is worth more than the next three weeks of tuning.
You watch five numbers, in this order, every morning. Not ten dashboards, not the platform recommendation panel, not the agency-style funnel chart. Five numbers, daily, in the same order, so you build pattern recognition over weeks. The number-one mistake non-marketers make is reading the wrong metric at the wrong stage: judging cost per click on day one with no conversions to compare against, or judging conversion rate on day three with a sample size of two. The right reading order matches the natural funnel: spend, clicks, click-through rate, conversions, cost per conversion. Your AI marketing employee should post these five every morning before you open the platform, so the narrative is in front of you when you look.
Is the account pacing inside your daily cap, or has the platform pushed you over again? Catch this first.
Are people clicking at all? Zero clicks for two days means the creative or audience is wrong, not the bid.
Tells you if the creative is resonating with the audience. Under 1% on search or 0.7% on social is a creative problem.
The only metric that pays the bills. Track them, even if you have to define a soft conversion to get early signal.
Compare against the gross margin of one customer. Below that line you are profitable, above it you are not.
A clean weekly routine for a solo founder running paid ads with an AI marketing employee fits inside ninety minutes total per week, split across five short blocks. The point of a routine is not discipline for its own sake. It is to make sure no creative ever runs for four days without a human glance, and no winner ever runs for a week without a budget bump. Without a routine, paid ads become an account you forget about until the bill arrives, the exact failure mode behind the founder ad-burn number earlier. With the routine, paid ads become a small predictable function in your week, more like checking inventory than gambling.
A few hundred euros across thirty days is enough to learn whether one creative-audience pair can convert. Less and the data is too thin to read. More than a thousand in month one without prior signal is gambling, not testing.
Search advertising on Google is usually the safest first platform because the intent is high and the feedback loop is fast. Social ads on Meta or LinkedIn reward better creative and audience targeting, which is harder to do well without a marketer on day one.
For volume work yes, for judgment calls no. A Sistava AI marketing employee handles drafting, reporting, and creative iteration well. Pricing, scaling, and brand positioning still need you. Realistic split: AI does eighty percent of hours, you do twenty percent of decisions.
Cost per conversion sits below the gross margin of one customer for three days running, or click-through rate stays above the platform benchmark with a steady flow of conversions. One good day is noise. Three days is signal.
When monthly ad spend crosses roughly three times your AI employee cost and you already have a profitable creative pattern, a part-time freelance media buyer pays back fast. Until then, the AI marketing employee plus a weekly routine is the right shape.
If the marketing function feels bigger than just paid ads (because for most solo founders it is, with content, email, SEO, and social all competing for time), the next read is the practical playbook for staffing the whole marketing function with AI Employees. It covers which roles to hire first, what each one owns in week one, and where the human still has to step in. Use it as the wider system that paid ads plugs into.
The honest framing for running paid ads without a marketer: the first month is tuition, not revenue. You are paying the platform to teach you which creative works on your audience, which audience converts at a price you can afford, and which offer on the landing page closes the loop. Expecting profit on day one breaks most founders. Going in with a small daily cap, a single platform, an AI marketing employee on creative and reporting, and a ninety-minute weekly routine is the shape that compounds. By month two or three, the same budget that felt like risk now feels like a small predictable line item. What changed is that you can finally read the data, judge the creative, and act on the winner inside the same week. That is what running paid ads without a marketer looks like in practice.