TL;DR. FastStrat ships three products: AI BrandOS generates your annual plan via a 58-question conversational workflow; StratMate is a 24/7 AI marketing consultant for refinement, scenario modeling, and approvals; Growth Engine executes the plan via Meta and Google integrations with real-time optimization. Six named agents (Martha, Brenda, Matt, Rikki, Dana, Pablo) power the work under a StratMate Manager orchestration layer. This post is the technical walkthrough of what each agent does, where each product fits, and how the pieces connect.
Most AI marketing vendors describe their product as a black box: you put a prompt in, marketing comes out. That abstraction makes for easy demos and impossible debugging. When something is wrong with your plan, you have no idea which part of the system got confused. When something is right, you have no idea why.
FastStrat is designed the other way. Each product is a clearly bounded workflow. Each agent owns a clearly bounded role. When you read an output, you can see which agent produced which part and why. This post is the walkthrough.
If you are still at the “should I use AI for marketing at all” layer, start with the AI marketing playbook for SMBs and the agency vs DIY vs AI comparison. If you want the side-by-side against hiring a traditional agency, see 5 things FastStrat does in 60 minutes that take an agency 3 months. What follows assumes you want to understand the machinery.
The six named agents, before we go into products
Three products draw on the same six agents. Introducing the agents first makes the product sections readable.
Martha — the Marketing Agent
Martha owns strategy end to end. Positioning, target audiences, channel mix logic, goal-setting, quarter-by-quarter roadmap. Martha is the synthesis layer: she consumes the outputs of the other agents and produces the plan’s spine. If you ask “what should we do this quarter and why,” Martha’s reasoning is the answer.
Inside BrandOS, Martha leads Questions 0-8 (the Foundational block) and Questions 27-40 (the Sales block, co-led with Matt). Inside StratMate, Martha is the consultant persona you talk to when you want strategic refinement.
Brenda — the Brand Agent
Brenda owns brand identity, recognition, and emotional connection. Voice and tone guidelines, visual positioning cues, brand-story structure, emotional territory. When a piece of generated content sounds “off,” the fix is usually a Brenda-layer issue: the context the writing agents had about your brand was incomplete.
Inside BrandOS, Brenda leads parts of Questions 41-53 (the Marketing block, co-led with Matt). Brenda is the agent that enforces voice consistency across the content calendar.
Matt — the Media Agent
Matt owns channels. Paid and earned media selection, budget allocation across channels, content cadence, and channel-specific format decisions. Should you lead with Meta or Google? How much goes to organic social vs paid? Where does email fit? Matt’s models are built around channel economics (CAC, LTV, conversion rates by channel).
Inside BrandOS, Matt co-leads Sales and Marketing blocks with Martha and Brenda. Matt also owns the 12-month content calendar output.
Rikki — the Research Agent
Rikki owns research. Market sizing, competitor intelligence, industry trend analysis, customer insight synthesis. When the plan quotes a market-size figure or a competitor pricing point, Rikki is the agent that sourced it and is responsible for citing it.
Inside BrandOS, Rikki leads Questions 9-20 (the Research block). Rikki runs both retrieval (pulling from live sources during the session) and synthesis (turning raw data into usable intelligence).
Dana — the Data Agent
Dana owns measurement. KPI frameworks, performance analysis, sentiment tracking, segmentation, attribution. Dana is the agent that translates “we want to grow” into a measurable set of primary and leading indicators tied to business goals.
Inside BrandOS, Dana generates the KPI framework at the end of the session. Inside Growth Engine, Dana runs live against your connected Meta and Google accounts, producing performance reports and flagging anomalies.
Pablo — the Product Agent
Pablo owns product. Positioning, pricing logic, launch sequencing, and the bridge from product truth to marketing message. Pablo is the agent that catches when marketing claims outrun what the product actually does, and that helps you find the message angle your product genuinely delivers on.
Inside BrandOS, Pablo leads Questions 21-26 (the Product block). Pablo also reviews the persona output to make sure messaging matches what the product actually solves.
StratMate Manager Agents — the orchestrators
The six named agents do not coordinate themselves. Above them sits an orchestration layer called the StratMate Manager Agents. Think of this layer as the chief strategist plus project director: it decides which agent handles what, in what order, with what shared context, and it resolves conflicts when two agents produce incompatible recommendations.
This layer is invisible to the user but is the reason the outputs feel coherent instead of like six separate consultants who never spoke. More in the core architecture of agents in FastStrat and agents vs models.
Product 1: AI BrandOS
What problem it solves
The single most common SMB marketing failure is not having a written strategic plan. Not a slide deck from 2022. A current, coherent annual plan covering positioning, ICP, channels, budget, calendar, and KPIs. BrandOS exists because the gap between “I need a plan” and “I have a plan” is where most SMBs get stuck.
The traditional paths to a plan are: hire an agency (10-14 weeks, 20,000-50,000 USD), DIY with templates (often incomplete, rarely executed), or use general AI tools like ChatGPT (scattered outputs, no operational structure). BrandOS is the fourth path: a structured conversation that produces the full artifact in a single session.
How it works under the hood
BrandOS runs a 58-question strategic workflow organized in five blocks. Each block has a lead agent. The workflow is not a form; it is a conversation where agents ask follow-ups, explain why a question matters, and help you reason toward answers when you do not have one prepared.
| Block | Questions | Lead agent(s) | What gets captured |
|---|---|---|---|
| Foundational | Q0-Q8 | Martha | Business context, stage, revenue, goals, constraints |
| Research | Q9-Q20 | Rikki | Market size, industry trends, competitor landscape, customer insights |
| Product | Q21-Q26 | Pablo | Product positioning, pricing logic, differentiation, launch context |
| Sales | Q27-Q40 | Martha + Matt | ICP, buyer personas, sales process, channel origination of leads |
| Marketing | Q41-Q53 | Brenda + Matt | Brand voice, positioning territory, channel mix, content pillars, budget |
| Implementation (Growth Suite only) | Q54-Q58 | Matt + Dana | 12-month calendar, content production, KPI framework, measurement plan |
Research sources
During the Research block, Rikki pulls from live sources: public company filings, industry trade publications, competitor websites and review platforms, public pricing data, and cited third-party market reports. Every market claim that appears in the final plan has a source attached. Rikki does not paraphrase statistics without citation, which is a deliberate design decision to prevent hallucinated benchmarks.
Output
At the end of a BrandOS session, the system produces:
- Annual marketing plan (narrative plus structured sections)
- Executive summary
- SWOT analysis rooted in the competitive research
- Financial projections for marketing spend and expected pipeline
- ICP and 2-3 buyer personas
- 12-month content calendar (Growth Suite)
- First month of content (Growth Suite)
- KPI framework with primary and leading indicators
Which agents power it
All six. Martha synthesizes; Rikki researches; Pablo handles product; Matt and Brenda handle marketing and channels; Dana closes with the KPI framework. The StratMate Manager Agents orchestrate the hand-offs.
Typical session length
Median 55 minutes. Output generation adds 5-15 minutes.
Related reading
For the structural logic of what belongs in a small-business annual plan, the 6-phase annual marketing plan guide. For a real SMB case walking this path, the SAGA Audiovisual case study. For how FastStrat compares to general-purpose AI tools on the same work, ChatGPT vs Claude vs FastStrat.
Product 2: StratMate
What problem it solves
A plan that sits in a PDF is dead weight. The question founders ask next is: what do I do when reality changes? A competitor drops prices. A channel stops working. Budget gets cut. The plan needs to adapt without dragging everyone back into a 60-minute regeneration.
StratMate is the layer that keeps the plan alive between full regenerations. It is a 24/7 conversational AI marketing consultant grounded in your BrandOS plan. You talk to it the way you would talk to a senior strategist on retainer, except the retainer is included.
How it works under the hood
StratMate has five capabilities that together distinguish it from a generic chatbot.
- Executive summaries. Ask StratMate for a one-paragraph or one-page summary of any section of your plan. For board meetings, for investor updates, for internal alignment.
- Strategic refinement. Ask StratMate to update positioning, adjust ICP, tighten the value proposition, rework the channel mix. Changes propagate to the plan with your approval.
- Tactical orchestration. Ask StratMate to produce a specific tactic (a campaign brief, an email sequence, a landing-page outline) and it routes to the right agent (Brenda for voice, Matt for channel, Pablo for product message) while keeping the output consistent with the plan.
- Experimentation manager. Define a hypothesis, and StratMate structures the test, sets success criteria, and tracks the result against the plan’s KPIs.
- What-if scenario modeling. “What if we cut paid spend by 40% next quarter?” StratMate projects the impact on pipeline and proposes compensating moves.
Human-in-the-loop
StratMate is designed around approval gates. Nothing automated ships to the outside world without your explicit approval. The agent produces, you approve, the change propagates. This is the architectural choice that distinguishes StratMate from autonomous agent systems that can take actions in your accounts without supervision.
Which agents power it
StratMate Manager Agents orchestrate. The six named agents are the specialists StratMate routes to. When you ask a question about brand voice, Brenda answers through StratMate. When you ask about attribution, Dana answers. The conversational layer is the Manager Agent; the substantive expertise is the named agent behind the response.
Product 3: Growth Engine
What problem it solves
A plan and a consultant are worth something. A plan, a consultant, and an execution layer that ships campaigns to Meta and Google and measures them in real time are worth more. Growth Engine is the execution layer.
Most agentic platforms stop at “generate campaign brief” and hand off to a human. Growth Engine closes the loop: API-connected to Meta and Google, content calendar integrated, performance data flowing back into the plan, optimization suggestions appearing in StratMate.
How it works under the hood
Three subsystems.
- API integrations. OAuth connections to Meta Ads, Google Ads, and (on the roadmap) LinkedIn Ads and TikTok Ads. Creative uploads, audience targeting, budget pacing, and reporting all happen through the integrations, not through copy-paste.
- Unified dashboard. One view of performance across channels, with attribution tied to the KPI framework Dana built in BrandOS. Breaks down by channel, campaign, creative, and persona.
- Growth Agent (real-time optimization). A background agent watches live performance, flags anomalies (spend pacing off, CPC spiking, conversion rate dropping), and proposes adjustments. Adjustments do not auto-apply without human approval.
Which agents power it
Dana owns the data and measurement layer (the dashboard, the KPI tracking, the attribution logic). The Growth Agent is a specialist execution agent that operates on top of Dana’s data. Matt contributes channel-level strategy recalibration. The StratMate Manager Agents route approvals.
How the three products connect
The simplest way to see the system: BrandOS produces the plan. StratMate keeps the plan alive. Growth Engine executes against the plan and feeds performance back.
- A founder starts in BrandOS. Session ends with a complete plan.
- The plan flows into StratMate. The founder uses StratMate for refinement and tactical orchestration between full regenerations.
- Approved campaigns flow to Growth Engine. Meta and Google execution and measurement run through the integrations.
- Performance data flows from Growth Engine back to StratMate. StratMate proposes adjustments to the plan. The founder approves. BrandOS updates.
The cycle is continuous. The plan is a living document, not a PDF.
Design decisions worth naming
Four architectural choices deserve a note because they shape what the system can and cannot do.
Named agents, not a single “AI”
Most vendors hide the internals behind one brand. We chose to name the agents because every marketing team (human or AI) has specialists. When you read a plan and want to understand why the channel mix is what it is, you should be able to see “this came from Matt, here is Matt’s reasoning, here are the inputs Matt used.” Accountability at the role level builds trust.
Plan-anchored, not artifact-generated
The system is organized around one durable artifact (the annual plan) that every other artifact traces back to. Ad copy cites the positioning in the plan. Personas cite the research in the plan. KPI reports compare against the plan’s goals. Without the anchor, AI-generated marketing drifts into incoherence.
Research sources are cited, not paraphrased
Rikki does not invent benchmarks. If a statistic appears in the plan, it has a source. This is a deliberate constraint that sometimes means the plan says “we were unable to find a credible public figure for this” instead of fabricating one.
Human approval gates at action boundaries
The agents can recommend, propose, and structure. They do not ship ads, send emails, or move budget without human approval. This is a product choice, not a technical limitation. We think the autonomous-action category is still too early to be the default for SMB marketing.
What the system is not
Three things we are not:
- Not a creative studio. The agents produce strong first drafts of copy and structured content. They do not produce Cannes Lions-caliber brand film. If your category depends on high-concept creative, pair FastStrat with a specialist agency or creative freelancer for that narrow layer.
- Not a CRM or marketing automation platform. We integrate with where you run your stack (HubSpot, Mailchimp, Meta, Google). We do not replace the system of record.
- Not a PR or influencer platform. Earned media strategy is advised by Matt; execution happens through other tools.
Where this system fits in the SMB marketing maturity curve
Not every SMB is ready for a three-product agentic stack. The right entry point depends on where you are in the evolution from founder-as-marketer to autonomous marketing. Our framework on SMB marketing maturity maps the four stages and the signals that tell you which product fits your current stage.
Common questions
Do I have to use all three products?
No. Foundation customers use BrandOS plus a limited StratMate tier. Growth Suite customers get all three as they ship. Many SMBs start with BrandOS and add the others as they grow.
Can I use FastStrat alongside an existing agency?
Yes, and many do. FastStrat handles strategy, planning, personas, research, calendar, and KPIs. The agency handles specialist creative, PR, or complex B2B relationship work. See the 60-minutes-vs-3-months comparison for the split.
Which agent is “the best”?
Wrong question. The agents are role specialists. A good plan uses all six. Asking which agent is best is like asking whether a good marketing team needs a strategist or an analyst.
Can the agents learn from our past data?
Yes, with your uploads. Past plans, past performance data, CRM exports, and customer interviews all feed context to the agents during BrandOS sessions and StratMate conversations.
What happens if I disagree with an agent’s recommendation?
You override it. The agents record your override as context and adjust downstream outputs accordingly. Nothing gets forced on you.
How is this different from ChatGPT or Claude?
ChatGPT and Claude are general-purpose models. Powerful, but not structured around marketing workflows. FastStrat is an orchestrated multi-agent system built specifically around the annual marketing plan as the anchoring artifact. The head-to-head comparison walks through this in detail.
Where is the pricing?
Current plans and pricing are on the FastStrat pricing page.
Next step
Meet the agents on the AI team page. The FAQ covers the common product questions. For broader context on the shift to agentic marketing platforms, AI marketing trends for SMBs in 2026 and how much a small business should spend on marketing.
About the author. Walter Von Roestel is CEO of FastStrat and the product architect behind the BrandOS, StratMate, and Growth Engine system design.
