AI for Web Agencies: Sell More Websites (Without Cold Canvassing)
AI for web agencies: find companies with an outdated or slow website, contact them with a concrete observation, and let AI handle research and follow-up.
Table of contents
How can web agencies sell more websites with AI?
By stopping calling blind. A web agency has something most other industries lack: the customer's own website is public, and it can be measured. Outdated software, slow loading on mobile, missing tracking: these are concrete, visible problems that you can approach them about. AI can take the time-consuming part of that: screen many sites, write a short report per company, produce a draft of the first contact and remember the follow-up. The salesperson reads, corrects, calls and takes the conversation.
TL;DR:
- Selling websites is hard because everyone has one. What works is timing and relevance: contacting companies that demonstrably have a problem.
- Public tools can show technical obsolescence, slow pages and missing tracking. AI gathers the results into a short report the salesperson can use.
- The person calls and judges. AI prepares and follows up. Anything that goes to a customer, and anything with a price in it, is approved by a human.
- Keep to companies and relevant decision-makers, and clarify the rules on calls and marketing before you start.
Why is it hard to sell websites?
It is rarely competition alone that makes it hard. The market is characterised by website builders, which let a company make a website itself, and by a large number of freelancers and small agencies. Most companies already have a website, and many are satisfied enough not to want to change it.
That means the big problem is timing. You call companies that do not know they have a problem. Or companies that have just had a new website made. The conversion on cold canvassing, where you ring down a list, is typically low, and it is tiring work for a salesperson.
At the same time, the agency's salespeople spend much of their time on something other than selling: finding leads, researching them, writing quotes, following up. Studies of salespeople's time use, including Salesforce's State of Sales, show that only a small part of the working week goes to actual selling. For web agencies it is often worse, because each quote requires a technical assessment.
What is "sniper selling", and why does it suit web agencies?
The shotgun approach is to call a hundred companies in the hope that two are ready for a new website right now. The sniper uses data to find the ones that demonstrably have a problem, and contacts them with a concrete observation instead of a generic pitch.
It works because an observation with data behind it is something quite different from a sales call. "I am calling because your website is slow on mobile, and it probably costs you visitors" is a service, not a pitch, if it is accurate and you have checked it.
Three types of signal are worth looking for:
1. Technical obsolescence. A website running an old version of its publishing system or on discontinued extensions is a sign that maintenance is not a priority. Tools such as BuiltWith and Wappalyzer can reveal which technologies a site uses.
2. Performance and user experience. Google measures websites on so-called Core Web Vitals: how quickly the main content loads, how quickly the page responds, and whether the layout jumps around. Google's recommended thresholds are that the main content loads within 2.5 seconds, that the page responds in under 200 milliseconds, and that the layout is stable. Google PageSpeed Insights shows how a page performs. A page that is slow on mobile is a concrete, measurable pain that you can show with a screenshot.
3. Missing measurement. Companies without tracking of visits and enquiries are blind to their own results. They do not know whether the website works, and so cannot see what a new one would deliver. That is an opening for the conversation about what the website is supposed to achieve.
Be careful with three things. Always check the observation manually before you use it: an automatic tool can be wrong. Exaggerations quickly damage credibility. And remember that a company is not necessarily interested just because its website is old.
From data to conversation: how the flow can look
- Define the target group. Industry, region, size. Write it down. A system cannot prioritise on an idea that exists only in the head of the agency's best salesperson.
- Screen the sites automatically. Use public tools to fetch technology, speed and tracking for each company on the list.
- Make a short report per company. AI can summarise the results in a few lines and write a draft opening: what was measured, what it means, and what could be done about it.
- The salesperson reviews. Read the report, correct the draft, remove those that do not make sense, and contact the rest.
- After the conversation. Minutes, CRM fields and next steps are ready as suggestions. The salesperson approves.
- Follow-up. If the answer is "not now", the company is put in a follow-up plan, for example a new measurement in a month, and the salesperson is notified when the time has come.
Most of it is reading, summarising and writing, which is what AI is good at. What requires judgement is human all the way through.
What can AI concretely take in a web agency's sales?
Research and brief. Gather what you know, and what is publicly available, about a company into a short brief, so the salesperson starts prepared.
Quote drafts. A quote draft with your rates, where missing information is marked. The salesperson sets the price. We can build this kind of thing: see the quote example.
Follow-up on old enquiries. Many agencies have hundreds of earlier enquiries and quotes that were never followed up. A system can go through the archive, sort by a rule you can read yourself, and make drafts that the salesperson approves. It is a conversation you have already had, being resumed. See the example.
After the meeting. The conversation is transcribed, AI suggests the outcome and writes an email draft. We have built a sales cockpit for exactly that. See also the CRM example.
Tasks for the project. When the agreement is closed, the promises from the sales meeting can become tasks with an owner, so nothing is lost in the handover to the project team. See the meeting example.
How do you calculate it?
The most important question is: what does a booked meeting cost, and what is a meeting worth? Set it up with your own figures:
| Measurement | What it tells you |
|---|---|
| Cost per month of sales (salary, tools, management time) | What do you spend today? |
| Number of meetings per month | What do you get out of it? |
| Share of meetings that become quotes, and share of quotes that become agreements | Where are the agreements lost? |
| Average project value and margin | What is an agreement worth? |
| Time per quote | How much can a draft save? |
Then you can see what one extra agreement per quarter would mean, and how much time must be freed up for a solution to be worth building. The price of a solution depends on how many systems need to be connected and what it is to do on its own. Agent360 prices according to the task.
How do you handle the typical objections?
Three classics in web agency sales, and how data can change the answer:
"We already have a web person." Do not argue on price. Show a concrete observation: "Your site is slow on mobile. Would you like to see the report?" You win on relevance and timing, not by saying that you are better.
"It is too expensive." Ask what the website is supposed to achieve, and show what the current solution costs in lost enquiries. Go through your lost quotes: which objections recur, and what worked when you won? A system can help find the patterns in your own conversations.
"We are not ready yet." That is not a no. Agree when you will follow up, and what you will send (for example a new measurement in a month). The follow-up can be ready as a draft, but a person sends it.
Which tools are involved?
There is no single product that does it all. Typically a mix is involved:
- Site screening: public tools to see technology, speed and tracking (e.g. BuiltWith, Wappalyzer and Google PageSpeed Insights).
- Contact data: your own contacts and public sources such as the CVR. Use only company and role information you have a basis for.
- CRM: whatever you already use, e.g. HubSpot, Pipedrive or Salesforce.
- Calendar and booking.
- Automation: tools to connect systems, if the connection is not built in.
- AI: to summarise, write drafts and find patterns.
It is not the number of tools that decides the result, but whether data moves from one layer to the next without anyone having to type it: from screening to brief, from conversation to CRM, and from CRM to follow-up. That connection is what we build.
Which rules do you need to have under control?
This is not legal advice, but these are the questions you must be able to answer:
- GDPR. Contact persons at companies are persons. You must have a clear basis for processing their information, inform them, and be able to delete it if they ask you to.
- The Danish Marketing Practices Act. Rules on unsolicited approaches and on whom you may contact in which way. The rules differ for consumers and companies, and for telephone, email and SMS. For consumers there is also the Robinson list.
- AI voice calls. This is a separate decision with legal and relationship questions. Our partner JesperAI works with voice agents. Get a legal assessment of your specific flow before you go that way. The alternative is that AI prepares and a person calls.
- Public data. Measuring a public website is not in itself a problem, but a list of persons is.
The biggest mistakes
Starting in the wrong place. Many begin with a chatbot on their own website. That can be fine, but it converts incoming traffic. It does not create new conversations.
Automating without data. A random list gives random results. Screen first, and contact only those with a real problem.
Giving up too early. Most decisions take several touches. Plan a reasonable follow-up, and stop when the recipient says no.
Measuring the wrong thing. The number of quotes and revenue come at the end. Measure along the way: the number of companies screened, contacts, meetings and quotes, so you can see where it goes wrong.
Forgetting the salespeople. Systems that salespeople do not trust get bypassed. Let them take part in defining what is worth automating.
How do you get started?
- Measure the current state for a month. Meetings, quotes, agreements and time per activity.
- Write down your target group. Industry, size, region and the signals that make a company interesting.
- Try the screening on a small scale. For example on a couple of hundred companies. How accurate is it when you check manually?
- Let AI make briefs and drafts, and let the salesperson correct. Measure what gets corrected.
- Measure again. Were there more meetings? Did less time go to preparation?
- Build on. Quotes, follow-up and CRM, one at a time.
How long this takes depends on your systems and your data. We will say that once we have seen them.
How Agent360 can help
Agent360 does not sell a finished sales system to web agencies. We build AI solutions in your own systems and call it Your AI Department:
- We think it through with you. We map where the time goes in your sales, and what is worth starting with.
- We build it with you. On top of the systems you have, e.g. CRM, email and calendar. You decide what the system may do on its own. Everything else waits for approval.
- We keep track of it. Tested, reviewed and in line with your rules for data.
If you are an owner or sales lead and want to understand for yourself what AI can be used for in your agency, we have a personal AI mentoring programme, where you learn on your own tasks, with a mentor at your side. You do not need to be able to code.
See more examples on the use-cases page.
Frequently asked questions about AI in web agency sales
Q: What does it cost to use AI in a web agency's sales? It depends on the task and on the systems that need to be connected. Agent360 prices according to the task. In the 20-minute free consultation we can say roughly what it requires, and whether it is worth building.
Q: Is it lawful to contact companies this way? It depends on whom you contact, how and why. The Danish Marketing Practices Act and GDPR apply, and the rules differ for telephone, email and SMS, and for consumers and companies. Have a lawyer review your specific flow.
Q: Can AI call prospects for us? Technically, voice agents exist, and our partner JesperAI works with them. That raises legal and relationship questions that should be clarified first. Many agencies choose to have AI prepare and a person call.
Q: How quickly can we see results? You can measure that if you have measured the current state first. We do not promise a time before we have seen your systems and data.
Q: Can an agency use AI without IT expertise? Yes, if someone builds and maintains the solution. That is exactly what we do, so you do not need to be able to do it yourselves.
Q: What happens to leads who say "not now"? They are put in a follow-up plan, for example a new measurement of the website in a month. The system can make the draft and remind the salesperson. The salesperson sends it.
Q: Can you build on top of our CRM? We build on the systems you have. Which connections are possible depends on the system and the provider. We contact your providers and help with access. See examples on the use-cases page.
Q: What is the first step? Measure what your sales cost and deliver today. Write down your target group. Test the screening on a small sample. Then choose the task that takes the most time and is the most uniform as the first to automate.
Conclusion
The market for websites is not saturated, but it is hard to reach with a generic approach. A web agency has the advantage that the customer's website can be measured. Use that to find the right companies, contact them with a concrete observation, and let AI take the preparation, reports, drafts and follow-up. The salespeople spend their time on the conversations.
Want to hear what we could build for your sales? Book 20 minutes of free consultation with Agent360. We listen to where the time goes, and say honestly what is worth building.
Sources
- Google, Core Web Vitals and PageSpeed Insights (web.dev)
- Salesforce, State of Sales (salespeople's time use)
- GDPR (Regulation 2016/679) and the Danish Marketing Practices Act
Contents
Should your sales reps spend their time talking?
We have built it for our own sales: the call is transcribed, the AI suggests the outcome, and the email is ready. We look at how your sales run today.
Book a reviewShould your sales reps spend their time talking?
We have built it for our own sales: the call is transcribed, the AI suggests the outcome, and the email is ready. We look at how your sales run today.
How we built it · Sales & follow-up
Sales cockpit with AI
A call list for salespeople. The call is transcribed in Danish, AI suggests the outcome (e.g. “Call again”) and drafts the email based on what the customer said.
- The salesperson calls
- AI writes transcript and email
- The salesperson sends
Related articles

AI in the Call Centre: How to Handle Peak Loads Without Hiring More People
AI in the call centre: handle peak loads, follow-up work and knowledge search without losing quality. Frameworks, rules and measures before you build.

AI for Car Sales & Leasing: Filling the Test-Drive Calendar
AI for car dealers and leasing companies: faster replies, follow-up on expiring leases and old leads, test-drive booking and service. Done with control.

AI for Events and Courses: Registration, Reminders and Delivery
How AI can support administration around events and courses: registrations, waiting lists, reminders, the instructor pack and the rules for invitations.