How an Adjacent engagement works.

Most agencies start with a discovery session. We start before it.

By the time we sit down with a prospective client, we have already run an AI-assisted research phase covering their competitive environment, relevant search landscape, positioning patterns in their sector, and observable gaps in how their category communicates. This takes between 24 and 48 hours depending on complexity. It produces an internal document we use to inform the conversation, not a slide deck we present back.

The effect is practical. We arrive at the first meeting with context, not questions. The conversation starts further along. Time is not spent assembling basic information we should have gathered in advance.

It also means we can tell within the first conversation whether we have something useful to offer, and we will say so if we do not. That is a more honest starting point than the alternative.

We use the first meeting to understand what the research cannot tell us: what the client's actual problem is, how they have tried to address it, what constraints are real and what are assumed, and what success genuinely needs to look like for them.

We do not pitch in this conversation. We ask questions and listen. We may challenge the framing of the brief if the research suggests the stated problem is not the real one. This is not confrontational; it is the most useful thing we can do early. Clients who have worked with traditional agencies sometimes find it unusual. A brief that is accepted uncritically tends to produce work that solves the stated problem rather than the actual one.

We follow up within 48 hours with a short document: what we think the actual problem is, what an Adjacent approach to it would look like, and an honest assessment of whether we are the right fit. If we are not, we say so.

Once an engagement is confirmed, we run a fuller version of the research phase with access to the client's existing data, analytics, and any internal material they can share.

What this produces for digital marketing and SEO engagements

A comprehensive map of the search landscape: competitor keyword rankings, content coverage by topic cluster, search volume and difficulty data for relevant keyword sets, identification of high-intent queries that competitors have missed or underserved, and a read on which existing client pages are close to performing and which are not worth preserving.

What this produces for web and messaging engagements

A competitive messaging analysis: the verbatim language patterns used across comparable organisations, the claims that are oversaturated in the category, the positioning angles that remain unclaimed, and a map of the available differentiation territory.

What this produces for recruitment engagements

A market map of candidates relevant to the brief, drawn from LinkedIn, portfolio platforms, industry publications, award shortlists, and conference speaker databases. Coverage that manual research would require three weeks to match.

Strategy at Adjacent starts from the research base and is built outward from there. The goal is a recommendation we have already argued against before we present it.

AI is used during this phase to generate alternative strategic framings and surface counterarguments to the position we are developing. We use it as a critical reader, not a co-author. If our recommendation does not hold up against the strongest version of the opposing case, we revise it. If it does, we can explain why with evidence.

For digital marketing and SEO engagements, this phase produces a content architecture and prioritisation framework grounded in what is achievable and winnable, not just what looks attractive in a chart. We distinguish between traffic and pipeline intent. Both have a place; the balance depends on what the business actually needs.

For web and messaging engagements, this phase produces a messaging architecture: the specific claims to make, the language to use, the things not to say because the category already saturates them, and the structural decisions about how the website should be organised.

The strategy document is short and specific. It is a set of decisions with the reasoning behind them, not a presentation designed to be impressive. Clients should be able to read it and immediately understand what we are going to do and why.

Production varies by service type. The structural principle is consistent: AI handles the work that does not require human judgment, and senior people handle the work that does.

Content programmes

AI handles initial research for each piece, structural outlines based on search intent and brief, and first-pass drafts. Senior editors review against the brief, rewrite for voice and accuracy, apply quality control, and sign off before anything goes to the client. The output is content produced at a pace a traditional editorial team cannot match, at a quality standard that requires human judgment to reach.

The ratio of AI contribution to human editing varies by complexity. A straightforward SEO article might involve 40% human editing. A piece requiring original argument, expert framing, or nuanced positioning will be closer to 80%. There is no fixed rule. The standard is that the piece has to be good, and a senior editor determines whether it is.

Web and messaging projects

Copy is drafted against the messaging architecture established in strategy, then edited for voice, precision, and accuracy. Design and development follow the messaging brief, not the other way around. Every page has a defined conversion purpose and is structured around it.

Recruitment

The AI-assembled long-list is reviewed by a senior consultant against the specific brief. Work quality is assessed independently. Candidates who match the data profile but would clearly not fit the specific context are removed. The shortlist presented to the client is small and considered; typically four to six candidates, not twelve. Every candidate on it has been assessed by a person, not just matched by an algorithm.

Work is delivered as it is completed, with a structured review window. We do not hold deliverables pending an open-ended feedback cycle. Clients receive work, review it in a defined window, and we act on consolidated feedback. This keeps projects moving and prevents the scope drift that typically extends timelines past the point where they are useful.

Documentation is standard and complete. Brand voice frameworks, content architectures, SEO strategies, and messaging rationales are delivered in formats the client's team can use independently. The objective is that when the engagement ends, the client understands what has been done and why, and can continue without us if they choose to.

Timelines stated at the outset are commitments. If something on our side changes, we say so immediately. If something on the client's side changes the scope, we have that conversation early rather than absorbing it silently and delivering late.

Project engagements have a defined scope, timeline, and end point. They suit clients with a specific problem to solve or a specific output to produce. The engagement closes when the deliverables are complete and documented.

Ongoing retainers suit clients running programmes that require consistent output: content at volume, SEO programmes that need regular iteration, or recruitment needs that recur. Retainers are scoped monthly and reviewed quarterly. They can be adjusted as priorities shift. We do not lock clients into structures that no longer serve them.

Most clients begin with a project and move to a retainer once the strategic foundation is in place. Some begin with a retainer and define scope as they go. Neither is preferable in principle. The question is which structure actually fits what the business needs to do.

One distinction that matters in practice: a retainer is not a subscription for deliverables. It is a structure for ongoing strategic work, with output that compounds over time. A content retainer at month six should be producing results that a project at month one could not, because the understanding of the audience, the keyword architecture, and the performance data are all more developed. That compounding is part of what makes ongoing work worth the commitment.