Automation is good at doing the same thing the same way every time. But a lot of the work in a growing agency isn't that predictable, a support ticket might need three different follow-up questions before it can be resolved, a research task might need to chase down whichever source turns out to be relevant. That's the gap agents fill: work where the right next step depends on what's found along the way.
What makes it an agent, not automation
The distinction matters because it changes what you can hand off. A simple automation runs a fixed sequence, if this, then that. An agent is given a goal and a set of tools, and it works out the steps itself:
- It reads and interprets: an inbound message, a document, a dataset, rather than matching a fixed pattern.
- It decides what to do next based on what it finds, not a pre-written script.
- It handles multi-step tasks: gather, check, act, follow up, without a human triggering each step.
- It knows its own limits and hands off to a person when a decision needs one.
An automation does what you told it to. An agent does what you need done, and tells you when it's not sure.
Where agents earn their keep
The clearest wins we see in UK agencies and service firms:
- Customer support: triaging incoming queries, resolving the straightforward ones, and escalating the rest with full context attached.
- Research: pulling together competitor intelligence, prospect background, or market data on demand instead of a person spending an afternoon on it.
- Data-gathering: monitoring sources, compiling updates, and flagging what's actually changed rather than a person checking manually.
Retainer model, typically £379–£2,000/month. This covers the agent's operation, monitoring and ongoing tuning as your business changes. Initial build scope is agreed and quoted before we start.
Why a retainer, not a one-off build
Agents are only as good as the boundaries and knowledge they're given, and both drift as your business changes, new products, new edge cases, new tools. A retainer means someone is watching how the agent performs and adjusting it, rather than shipping it once and hoping it still fits in six months.
Frequently asked questions
What's the difference between an agent and simple automation?
Automation follows a fixed set of steps. An agent handles tasks where the right next step depends on what it finds, it can read, decide, and act across multiple steps rather than following one script.
What tasks can an autonomous agent handle?
Common uses are customer support triage and first-line replies, research and data-gathering across the web or your own systems, and monitoring tasks that need judgement rather than a fixed rule.
How is this priced?
As a monthly retainer, typically £379 to £2,000, covering the agent's operation, monitoring and ongoing tuning. Initial build scope is agreed before we start.
Does the agent work unsupervised?
Agents are given clear boundaries on what they can decide versus what they must hand to a person. Where a decision has real consequences, the agent flags it rather than acting alone.
Related services
- Ishigai vs a £5,000 consultancy audit: what the price difference buys.
- Ishigai vs building it yourself: when Zapier or n8n is the better answer.
- Ishigai vs doing nothing: when waiting is the right call.
- AI Opportunity Audit: a fixed-price, ranked and costed map of where AI saves your firm hours. From £750.
- Process Automation: repetitive admin turned into monitored pipelines. Fixed price.
- AI Security Audit: hands-on testing of your chatbots and agents for prompt injection and data leaks.
- Worked example: the full audit deliverable for an 18-person recruitment agency.
The weekly note. One short email a week from Akashuin, our research blog: what UK firms are actually doing with AI, the numbers behind it, and one thing you can try that week. Free, and one click to unsubscribe.