How to Use Actus Agent for Competitive Research Without Manual Tracking

How to Use Actus Agent for Competitive Research Without Manual Tracking

Competitive research is one of those tasks everyone agrees is important and almost no one does consistently. You know you should check what competitors are publishing, how their pricing has changed, whether they launched a new service, and what their customers are saying. Then a client emergency lands, the week disappears, and the research never happens.

The cost is not dramatic in any single week. It compounds. You walk into a sales call without knowing a competitor just cut prices. You publish content that overlaps a piece they published three days ago. You miss a positioning shift that would have changed your offer.

Actus Agent can run competitive research as a recurring workflow: visit the same sources on a schedule, extract what changed, and deliver a short brief you can actually read. This guide shows how to set that up without turning it into a surveillance project or a pile of unread reports.

What Competitive Research Should Answer

Before you automate anything, decide the questions the research must answer. A useful brief is not a dump of every page a competitor published. It answers a small set of operator questions.

For most small businesses and agencies, those questions are:

• Did their offer, pricing, or packaging change?
• Did they launch a service, location, or campaign?
• What content are they publishing, and is it aimed at the same buyer as yours?
• Are customers complaining about something you could solve better?
• Is there a claim they make that you should address or avoid?

If a data point does not change a decision, do not collect it. Tracking follower counts, every social post, and every blog headline creates noise. Tracking offer pages, pricing pages, and recent case studies creates signal.

Choose a Small Competitor Set

Five competitors is enough. Ten is the upper limit before the brief becomes unreadable.

Split them into two groups:

Direct competitors sell the same service to the same buyer in the same market. For a Southwest Florida web studio, that might be other local agencies that build sites for contractors and service businesses.

Adjacent competitors solve the same problem with a different model. A template marketplace, a freelancer marketplace, or a national agency that targets the same buyer but not the same geography.

Write down, for each competitor:

• Company name
• Homepage URL
• Pricing or services URL, if it exists
• Blog or resources URL
• One public review source (Google, Clutch, G2, or a relevant directory)

That list is the input to the workflow. Do not ask the agent to "find my competitors" every week. You already know who matters. The agent's job is to notice change.

What to Monitor, and What to Ignore

A practical monitoring list:

Offer and pricing pages. New packages, removed services, price changes, new guarantees, new timelines.

Homepage messaging. Headline, primary call to action, proof (logos, reviews, case counts).

Recent content. The last three blog posts or resource pages. Title, date if visible, and the audience they seem written for.

Public reviews. New reviews that mention a specific strength or complaint. Not the star average by itself.

Careers or "now hiring" pages, if relevant. A sudden hiring push in a service line can signal where they are investing.

Ignore:

• Social engagement metrics unless social is your actual channel
• Design tweaks that do not change the offer
• Press mentions that do not affect your buyer
• Anything behind a login

Actus Agent should only use information that is publicly available. Do not instruct it to bypass logins, scrape private communities, or collect personal data that is not needed for a business comparison.

Build the Weekly Brief Workflow

Here is a workflow you can hand to Actus Agent.

Goal: Every Monday morning, produce a one-page competitive brief covering five named competitors.

Steps:

1. Open each competitor's homepage, services or pricing page, and blog or resources page.
2. Compare what you see to the notes from the previous run. If this is the first run, record a baseline instead of claiming something "changed."
3. For each competitor, write at most five bullets:
   • Offer or pricing change, or "no visible change"
   • Homepage message, in one sentence
   • Up to two recent content titles and who they appear to target
   • One notable review theme, only if a new review is visible
   • One implication for our business, phrased as a question or a possible action
4. End with a "Worth a look" section: at most three items that a human should open personally.
5. Do not invent prices, dates, or quotes. If a price is not on the page, write "pricing not published."
6. Save the brief where the team already reads things: email, a doc, or a Slack channel.

The instruction "do not invent" matters. Competitive briefs lose trust the first time they cite a price that was never on the page.

How to Write the First Baseline

The first run should not look for changes. It should create the baseline the later runs compare against.

Ask for a baseline card per competitor:

• What they sell, in one sentence
• Who they appear to serve
• How they ask visitors to convert (call, form, calendar, checkout)
• Whether pricing is public, and the published range if it is
• Two proof points they lead with (reviews, case studies, years, guarantees)
• Gaps you notice relative to your own offer

Store that baseline. On week two, the workflow compares against it. On week six, you still have a record of what the market looked like when you started.

A baseline is also useful in sales. When a prospect says "we are also talking to X," you can speak to X's actual public offer instead of a vague impression.

Turn Findings Into Decisions, Not Trivia

A brief that says "Competitor A published three blog posts" is trivia. A brief that says "Competitor A is now leading with a 14-day launch promise on contractor sites, and their newest posts target HVAC owners" is a decision input.

Train the workflow to attach one implication to each real change. Implications should be questions, not orders:

• Should we answer this claim on our own site?
• Is this a segment we are ignoring?
• Does our proposal need a clearer timeline because they are selling speed?
• Is their new package overlapping the offer we just launched?

You still make the decision. The agent surfaces the question while the information is fresh.

A Worked Example

Imagine a small agency that builds websites for local service businesses. The Monday brief might look like this.

Competitor: Harbor & Co.

• Pricing page now lists a "starter site" and a "growth site." The starter price is published; the growth price says "custom." No prices were public last month.
• Homepage headline shifted from "beautiful websites" to "websites that book jobs."
• Two new posts target roofing companies and med spas.
• Implication: they are moving toward the same buyer and publishing price on the entry offer. Do we need a public starting price so we are not the only opaque option?

Competitor: Northline Studio

• No visible change to services or pricing.
• New case study: a painter in Naples, with a quoted increase in form submissions. Treat the number as their claim, not a verified result.
• Implication: they are using local proof. Do we have a similar story we have not published?

Worth a look

• Harbor & Co. pricing page
• Northline case study

That is a brief a founder can read in four minutes. It does not pretend to know Harbor's close rate. It tells you what changed in public and what to decide.

How Often to Run It

Weekly is right if competitors publish often or you are in an active sales cycle. Biweekly is enough for stable local markets. Monthly is too slow if you use the research in live deals.

Match the depth to the cadence:

• Weekly: changes only, plus the baseline on file
• Monthly: a slightly deeper pass on messaging and content themes
• Quarterly: revisit whether the competitor set is still the right five

Do not run a deep audit every Monday. You will stop reading it.

Keep a Human Review on Claims

Competitive research is a good candidate for near-full autonomy because the output is internal. You are not emailing the competitor. You are reading a brief.

Still review it before you repeat a claim in a sales call or on your website. Specifically check:

• Prices and package names
• Any customer result the brief quotes
• Whether "no change" might mean the page failed to load

If the agent could not open a page, the brief should say so. Silence is not the same as "nothing changed."

What This Workflow Does Not Do

It does not:

• Predict a competitor's revenue
• Tell you their ad spend unless that number is published
• Replace talking to your own customers
• Justify copying their positioning

The point is awareness. You should know the public market you are selling into. You should not build your strategy as a reaction to every post they publish.

Common Failure Modes

The brief is too long. Cap it. Five competitors, five bullets, three links. If you need more, that is a separate deep dive, not the weekly brief.

The agent treats marketing copy as fact. Instruct it to label claims. "They say projects launch in 14 days" is accurate. "They launch projects in 14 days" is not something you know.

The competitor list goes stale. Once a quarter, ask whether one of the five should be replaced. Markets shift. A freelancer who was irrelevant in January may be your most common alternative by June.

You collect research and never use it. Put the brief in the same place proposals get written. If it lives in a folder no one opens, the workflow is theater.

How Actus Agent Fits

This is a natural Actus Agent workflow because it is recurring, multi-step, and judgment-based. The agent opens real pages, compares them to a stored baseline, and writes a short brief in your language. You are not building a brittle scraper for every site layout.

You still own the competitor list, the questions, and the decisions. The agent owns the Monday morning pass you would otherwise skip.

Start with five competitors and a baseline run. Schedule the comparison for the following week. Read it. If the brief changes what you say in a proposal or on your site, keep it. If it does not, narrow the questions until it does.

Learn more at actusagent.cc.