How to involve your agencies – and how to measure whether any of this is working
The instinct when the estate changes shape is to re-tender your agencies. Sometimes that’s right. Sometimes it destroys the thing you need most: continuity.
Over the past two weeks I’ve written about why portfolios grow and what the advantages actually are, and how to plan when nobody gave you a plan. This week is the external lens: how to think about your agency relationships as the estate grows, and how to measure whether any of this is working.
If you’ve done the work from the last two weeks – tiered your sites, defined what success looks like, built the governance model – you’re in a strong position. You’re not going to your agency with a vague request. You’re going with a portfolio strategy that needs the right partner to execute it.
Making a clear-eyed decision about your agency
Some agencies are built for depth on one brand. Others are built to scale across many. The skills are different. A team that’s excellent at deep, bespoke work on a single flagship may struggle with the context-switching and pattern recognition that a multi-site estate demands.
If your current agency was hired for a single-site engagement and your estate has tripled, it’s worth asking whether the fit has changed. Not as a criticism – as a recognition that the job has changed. The agency that got you to this point deserves the conversation about whether they can grow with you.
The honest question is whether you’re re-tendering because the agency genuinely can’t scale with you, or because nobody’s had the conversation about what the new shape of the relationship should look like. Often it’s the second one. And that conversation is yours to start.
Every agency change costs six to twelve months of context. We’ve worked with a retailer running fifteen sites on fifteen different systems – none of them integrated. The ramp-up for a new partner would consume the first year. If the core relationship is sound but the scope has outgrown the original engagement, “how do we restructure this?” is almost always a better conversation than “let’s go to market.”
How to brief for a portfolio (and why it filters for the right partner)
Most agency briefs are written for a single site. “Build us a new checkout.” “Redesign the homepage.” “Migrate to a new platform.” When the estate is five, ten, twenty sites, single-site briefs create single-site thinking.
Brief for the portfolio. What are the tiers? What does each tier need? Where do you want shared components and where do you need distinct experiences? What does the team model look like and where does the agency sit within it?
The best briefs we’ve seen describe the estate as a system, not a collection of individual projects. They’re clear about which sites are flagship, which are operational, and which are experimental – and they ask the agency to propose how their approach flexes across those tiers.
Here’s the thing: a portfolio-level brief is the best filter you have. An agency that can’t answer it probably isn’t built for multi-site work. One that can will show you how they think about scale, not just how they execute on a single engagement. You’ll learn more about a potential partner from their response to a portfolio brief than from any credentials presentation.
Measuring what matters (and reporting it upward)
You can’t judge a £500k niche site by the flagship’s KPIs. Revenue per site is the obvious metric, but it tells you very little about whether the portfolio is healthy.
Better questions per tier:
For flagship sites – are you growing revenue and improving experience simultaneously? Is the investment producing compound returns or diminishing ones?
For operational sites – are they stable? Is the cost of maintaining them decreasing over time as shared infrastructure matures? Are they generating data and learnings that benefit the wider estate?
For experimental sites – are you learning fast enough? Is the investment producing insight that shapes decisions elsewhere? Are you comfortable killing experiments that don’t work?
The portfolio-level metric that matters most is whether each addition makes the estate stronger. The retailer that doubled their business through COVID acquisitions could measure this clearly – the third acquisition was cheaper to integrate than the first because the patterns existed. If site ten costs more per-site than site five did, something structural is wrong. If it costs less, the foundations are working.
This is also how you report upward. Boards don’t want site-by-site revenue tables that make the smaller brands look like underperformers. They want to know the estate is compounding. Show them the cost-per-integration trend. Show them the shared infrastructure ROI. Show them that the playbook you built is making each addition faster and cheaper. That’s a story executives understand, and it positions you as the person who owns the portfolio strategy – not just the person who keeps the sites running.
The compound
This is what it comes down to. A well-structured multi-site estate compounds. Shared infrastructure gets cheaper per site. Shared learnings accelerate decisions. The playbook you build for site three makes site ten faster and cheaper. The team – internal and external – gets better at the pattern.
The difference between an estate that compounds and one that drags is rarely talent or technology. It’s whether someone sat down and wrote the playbook – tiering, team model, governance, measurement – before the weight of the portfolio made it impossible to think clearly.
If you’ve followed this series, you already have the framework. Tier the sites. Shape the team. Build the governance. Brief your agencies for the portfolio. Measure what compounds, not just what converts.
The team that does this doesn’t just manage a growing estate. They’re the ones shaping it.