How our search for new customers led us to build Adapt

It started with the work Grant and I were doing

Grant and I created Adapt because we wanted a better way to find new customers for DOJO.

That is the starting point. We work with businesses that need to take card payments, and we wanted technology to help us find the right ones, understand their circumstances and keep a useful conversation moving. There is a lot of work before a business owner ever agrees to look at a proposal.

You need to find the business. Check what it does. Establish whether you can serve it. Find a suitable person to speak to. Understand enough to make the first conversation relevant. Then remember what was said, who is following up and when.

Those are ordinary sales tasks. They become difficult when the information and the work are spread across different places.

We wanted to automate more of that preparation in our own business. As we built Adapt, the wider purpose became clear: other teams trying to find customers face much of the same work, even when they sell something completely different.

Adapt is launching soon. This is the story behind it, the problem we are building it to address, and why we think bringing research, outreach and CRM together matters.

Why payments made the problem so clear

For a business owner, a card machine is part of getting paid. For a payments consultant, finding an appropriate customer involves understanding how that business operates.

A restaurant taking payments at tables has different requirements from a shop serving a queue, a dental practice collecting payment after treatment, or a visitor attraction with several payment points. The practical questions include where payments happen, how they connect to other software, how funds arrive and what happens when something goes wrong.

DOJO offers in-person and remote payment products, alongside integrations with EPOS, the software businesses use to record sales and manage service. Its public industry pages cover hospitality, leisure and entertainment, dental practices and personal services. Those are examples of the businesses relevant to our work, rather than a complete acceptance list. Explore DOJO’s products and business types.

Business owners also have alternatives. Worldpay offers in-store, mobile and online payment services. SumUp supplies portable readers and standalone devices. Square combines payment acceptance with its software. Their products, commercial arrangements and suitability differ. Our job is to understand whether DOJO fits the business in front of us.

That takes more than finding a business with a card machine and offering a headline rate. A useful comparison considers the whole arrangement: the business’s payment mix, existing equipment, contract, service requirements and the work involved in changing provider.

The difficulty of comparing providers has also received regulatory attention. The Payment Systems Regulator’s card-acquiring market review led to measures intended to improve information and make it easier for businesses to compare and switch. That is market context, rather than evidence that every business should change provider. Read the PSR’s review and remedies.

For us, the opportunity starts with a conversation about how the business gets paid today. Software should help us arrive prepared for that conversation.

What happens when the businesses you can serve change?

Eligibility is part of prospecting in payments. You can find an apparently suitable business, only to discover that its activities fall outside a provider’s current underwriting criteria. Underwriting is the provider’s assessment of whether it can accept a merchant and on what terms.

In our experience, repeatedly working within the same familiar sectors meant consultants were often approaching similar audiences. The challenge was finding relevant opportunities beyond the businesses everyone already knew to call.

Recent changes to DOJO’s supported-business guidance have broadened the opportunities we can explore. That makes the quality of our prospecting information more valuable: we need to identify businesses in the relevant categories, understand what they actually do and check their suitability against current guidance.

For example, the newer sales guidance we reviewed includes consideration of training businesses, driving instructors and certain legal services. The older guide we had placed those areas outside its supported categories. These examples help explain why we want a database we can search beyond our familiar audiences. Individual applications still need assessment under the current criteria.

There is an important difference between a business worth researching and a merchant approved for payment services. A database category cannot make an underwriting decision. Particular activities, payment arrangements and the provider’s assessment still matter.

The practical lesson applies beyond payments. A business might introduce a new service, expand into another region or start serving a different type of customer. Its existing prospect list may no longer reflect the opportunity.

What would you need to know before approaching that new audience? Who belongs in it? Which existing customers should be excluded? What would make your service useful to them? Those are the questions we wanted Adapt to help us work through.

Buying data gives you a starting point

A business database can help you find companies by sector or location. A purchased list can be useful. The work begins when you decide which records deserve your attention.

Consider an illustrative payments prospect: a company record points to a restaurant group. Before making contact, a consultant may still need to establish whether it trades under another name, which locations are operating, who handles payments and whether someone on the team has already spoken to it.

A registered office is not necessarily a trading location. A named director is not necessarily the person choosing the payment provider. A listed email address does not, on its own, establish that it still reaches the right person.

These distinctions affect the quality of the conversation. If you cannot explain why a company belongs on your list, a personalised opening line will not fix the underlying targeting.

We wanted research to remain attached to the business record: what information we have, where it came from, when it was checked and what still needs confirming. That gives a colleague something to work from rather than another spreadsheet to interpret.

Freshness matters at the level of individual facts. A newly checked website does not make every contact field current. A public company update can suggest something to investigate, but it does not tell you that the owner wants to buy.

That is how we want to use better information: to prepare a more informed approach and ask a better question.

The gaps between LinkedIn, email, calls and CRM

The difficulty is not confined to finding a name. It continues after you have one.

LinkedIn can help you understand someone’s role and professional context. Email gives you a way to introduce a relevant idea. A phone conversation can clarify a problem much faster than guessing from a website. A CRM should help the team remember the discussion and agree who does what next.

Each has a purpose. When the work is disconnected, the person selling has to carry the context between them.

Cold email makes the need for relevance particularly clear. The recipient has not asked for your pitch, so you need a credible reason for approaching that business and a message that explains it. If they reply, that response should guide the conversation instead of being lost while another scheduled message carries on.

For a small team, the consequence can be straightforward: time spent rebuilding a colleague’s research, an unnecessary second approach, or a useful conversation left without an owner.

We wanted one platform to support that sequence of work. Find the business, understand it, prepare the approach, record the response and carry the conversation forward.

A better opening starts with the customer’s situation

Automation can help prepare a message. The message still needs a reason to exist.

In payments, an opening about how a business currently takes payments can reveal far more than an immediate pitch about switching. For example:

• “How do payments and your EPOS work together at the moment?”
• “What happens when the figures do not match at the end of the day?”
• “How does the team resolve that, and what other work does it interrupt?”
• “What would it change for you if those payments were recorded together?”

Those questions belong in a conversation where the answers guide what comes next. If the owner says the current arrangement works well, that is useful information too. You should not manufacture a problem to make your proposal fit.

This reflects the thinking behind SPIN Selling: understand the situation, explore a difficulty the customer recognises, establish its effect and discuss the value of improving it.

For Adapt, the implication is practical. Research and message preparation should help a seller understand the customer. When someone replies, what they actually say should inform the follow-up. A change of director, a new location or an industry label is context for research, not a substitute for that conversation.

What we are bringing together in Adapt

Adapt is being built around that complete journey. Its purpose is to bring UK business information, outreach and CRM into the same working environment.

Research the businesses you want to reach. Start with the sectors, locations and characteristics that matter to your offer. Examine the information available for each business and distinguish useful evidence from details that still need checking.

Prepare the approach with that context in view. The email, LinkedIn message or call preparation should relate to the business you have researched. Your team should be able to review the wording and understand why the prospect was selected.

Keep the conversation connected to the record. Replies, notes and follow-up should help the next person understand what has happened. The purpose of a CRM is partly to make sure that a relationship does not depend on one person remembering every detail.

Learn from the work. If an audience does not respond, examine the targeting and the approach. If conversations start but go nowhere, look at what the customer said and how the team followed up. Activity becomes useful when it helps you make a better decision.

The Adapt platform pages explain the tools in more detail. This article describes why we are building them together. The public launch is coming soon, and we will explain the available features and access arrangements as it opens.

Why agents are part of the plan

Once the information and work are connected, an assistant can help with more than a single piece of writing.

We are building towards being able to give a task in ordinary language: find businesses matching our customer profile, show the evidence behind a shortlist, prepare an approach and help organise the follow-up.

That requires more than asking a model to produce persuasive text. The assistant needs the relevant business information, the team’s existing activity and access to tools that perform defined actions.

Adapt’s agent approach divides the work between research, insights, message preparation, coordination and recommendations. Vela coordinates the work. The point is to keep the context available as a task moves between stages, with the person using Adapt able to review the evidence and decide what to do. Read how Adapt’s agents are designed to work.

The longer-term ambition is for more of Adapt to work headlessly. In plain English, that means using its data and supported actions from an assistant or another application, without having to open a separate screen for every task. The visual app remains useful when you want to inspect a record, edit a campaign or review the pipeline.

We want the technology to fit the way someone is working, whether that is in a sales workspace or a conversation with an assistant.

Where Grok Bot and an agent harness fit

Products such as Grok Bot show how assistants are developing beyond individual chat replies. Cursor’s documentation describes persistent bots that can undertake tasks using a cloud computer, connected tools, files and a browser. Read Cursor’s Grok Bot documentation.

You may hear the surrounding software called an agent harness. It manages the context available to the model, the tools it can call, the progress of a task and the points where a person needs to take over. The model helps decide what to do; the harness provides the working environment in which those decisions become actions.

This matters to Adapt because an assistant needs access to useful business information and supported sales tools to help with prospecting. Our ambition is for Adapt to provide that access through standard connections. Mentioning Grok Bot here explains the direction of the technology; we are not announcing a dedicated Grok Bot integration.

What MCP means in a sales workflow

MCP stands for Model Context Protocol. It is a standard way for an AI application to discover and call tools supplied by another system. A connected tool might search business records or read a saved list. The application makes a structured request and receives a result it can use in the conversation. See the official MCP architecture explanation.

For example, an illustrative task could be:

Find independent dental practices in Manchester. Compare them with this file of businesses we have already contacted. Show me the remaining records and prepare opening messages for review.

An assistant would need access to both the business search and that file, together with instructions about the comparison. MCP supplies the connection to supported tools; it does not supply missing information or decide that a company is eligible for DOJO.

Adapt already has an MCP connector as part of the platform architecture. It exposes research and supported sales operations through an authorised account. Which actions can run depends on the connection’s permissions and the requirements of each tool. Explore Adapt’s MCP connector.

That connection is a foundation for the headless ambition. It makes it possible to bring company research into a conversation with a compatible assistant, then build towards more connected workflows. It does not mean handing an assistant an unrestricted instruction to contact everyone it finds.

What would better preparation let your team do?

Imagine beginning a prospecting session with a shortlist you can explain. You know why each business is relevant, which facts need checking and whether your team has already been in touch. A draft is available beside the research. When a reply arrives, the conversation and the follow-up have somewhere to live.

For a payments consultant, that could mean spending more of the day discussing how a merchant gets paid and less of it reconstructing a prospect list. For another business, the discussion will be about its own service and its own customers. The value comes from being better prepared and making it easier to carry a useful conversation through.

That is what Grant and I want from Adapt in our own business. We want to find customers we can help, arrive with relevant information and follow up properly. We believe other businesses want that too.

Adapt is launching soon. If finding customers currently means moving between lists, profiles, inboxes and notes, tell us how your team works. We can discuss where Adapt could help and what you would want to see when it launches.

You still bring the judgement, the service and the relationship. We are building Adapt to make the work around those conversations easier to manage.