Revenue Diversification: A Prudent Defense in an Uncertain Future

We’re helping an SEO agency stand up a paid media practice. For them, the move is about growth. They have the scale, the client relationships, and the demand to support it, and adding paid is a logical next step toward a bigger, stickier book of business.

But watching that engagement unfold sent my thinking in a different direction. I work with other SEO-only shops where one question on the table is not just "how do we grow faster." But "what happens to us if AI changes how people find and choose what to buy?" For those firms, adding a second service line is not just for offense. It could serve as a dual-purpose defensive hedge.

That’s the case I want to make here. If you run a single-offering firm, you’re essentially a concentrated bet on one version of the future. The uncertainty AI has introduced makes that bet riskier than it was three years ago, and diversifying your revenue could be a prudent way to manage that risk. We tend to frame it only as a growth play when it is at least as much a defensive one.

Why now? Because the range of plausible futures has widened (I love scenario planning, so you can view this through that lens if you’d like!) A few forces are worth naming specifically, because "AI uncertainty" is too vague to plan against:

• The deliverable itself gets commoditized. Work that used to take a skilled team a week starts taking an afternoon.  What is the price your clients are now willing to pay comes under competitive pressure?

• Clients pull the work in-house. The same (ever smarter) tools that pressure your pricing also lower the bar for a client to do a passable version themselves.  Set aside whether the client should do it themselves, lol.  But what if a number of them decide to try?

• Demand migrates between channels. The channel where your clients' customers pay attention shifts, and the budget shifts with it.

Any one of these can erode a single-offering firm's core without much warning, and to be fair these three are always a threat (and opportunity).  The key thing now is the pace of change driven by AI. Diversifying is a way to place bets across more than one version of the future. If your read on the next 18 months is uncertain, holding a second, uncorrelated source of value is a sensible response. I think of it as buying optionality, not just chasing top line.

Start with an honest question: are you exposed?

Step one though is to take a beat to see if this is actually something you need to worry about. Before you treat diversification as a hedge, confirm you have something to hedge against.

Run your core solution or deliverable against the three forces above. Is your deliverable getting commoditized? Are your clients starting to bring the work in-house? Is demand migrating to channels you do not serve? If the honest answers are mostly no, the defensive case may not apply to you.  

A concrete example. A paid media shop whose ICP is roofers, HVAC installers, and other home-services firms under $20M in revenue probably doesn’t need to lose sleep over AI. Their clients' customers still find them through local search and referral, even direct mail, the buying decision is trust-driven, and AI does not change much about how a homeowner picks a contractor. There is little here to hedge.

Two caveats keep this honest. First, a growth motive can still justify diversifying even when the defensive one does not. That is a different calculation with a different risk profile, and it’s not the argument I am making here. Second, your exposure is not only about your service. It is also about your book of business, meaning which of your clients sit in industries AI is reshaping fastest. That portfolio question deserves its own treatment, and I will come back to it in a later piece.

The second gate: can your firm take the punch?

Say you are exposed and the hedge makes sense. There is still a threshold to clear, and it has nothing to do with ambition. It has to do with whether your organization can absorb the shock of a distraction.

I tend to put a rough floor around $3M in revenue, but the number is just a proxy. What it really stands for is resilience.  Do you have: 

• A senior leadership team (SLT) that is actually in place and functioning, 

• Financials you can trust

• Real budgets in place, and 

• Operations and controls locked down well enough to run without constant founder intervention. 

None of it has to be perfect. It does have to be sturdy.

Here is the let’s keep it real test I would apply. You haven’t earned the right to diversify until your core solution delivery capability can run a full quarter without the founder in the room. If it cannot, a new service line will not diversify your risk. It will increase it, because the resource every new venture consumes most is the founder's attention, and it consumes that attention at exactly the moment your core business needs adaptation to survive the same pressure you are hedging against.

This is also where I try to be honest with founders about themselves. Some owners I’ve seen are uniquely wired for the new and shiny, and someone in their orbit, a partner or a number two, is often working to keep them focused. I’m not telling you to chase the shiny thing. I am giving you the discipline test that decides whether you are allowed to. 

If your go-to-market, retention, leadership, and finance functions are a mess, trying to diversify by launching or buying a new channel is one of the more reliable ways to blow up the business you already have.

Diversify along your customer, not away from them

A common objection to any of this is the fear of becoming a "we do everything" shop that stands for nothing. It is a real risk, but it is misunderstood, IMO. Dilution is not a function of how many services you offer. It is a function of which direction you expand.

There are four basic ways to add a service, and they sort cleanly onto an identity-risk line:

• Solve a new or emerging problem for your existing clients.

• Solve a different, existing problem for your existing clients.

• Take your existing service to a new type of client.

• Build a new service for new clients.

The first two hold your customer constant and expand what you solve for them. That direction is accretive to your positioning. You end up owning more of a tightly defined customer's world, and you meet a real preference: clients inside a coherent niche often want one partner to consolidate around rather than managing three. The second two reach for customers you do not yet understand, and that is where the "we do everything" mush actually comes from. It is also where you could explore a separate brand or a distinct position.

So the rule is simpler than the fear suggests. Diversify along your customer and a broader offering strengthens your position. Diversify away from your customer and you take on identity risk.

An example of doing it well. I work with an outsourced bookkeeping and accounting firm that serves only creative and digital agencies. They recently added a small but growing M&A advisory practice. They did not broaden their client base to do it. They went deeper into the same one, moving from their clients' books to their clients' transactions. That is not dilution. It is deepening, and it makes the original relationship stickier rather than weaker.

One more distinction worth holding. Some adjacent services are additive, just another line on the rate card. Others are integrative, meaning they make everything else you do more valuable. Analytics and reporting, the GA4 and measurement layer, is the clearest example for folks in the agency space. It is connective tissue across paid, SEO, and content rather than a silo of its own. For other types of consulting or advisory services, imaging investing in a technology that allows your delivery team to provide even greater value services to the same ICP. When you can, reach for the integrative adjacency. It compounds.

Make sure it really is a hedge

Here is the trap. A second service feels like diversification, but if it carries the same risk as your core, you have not hedged anything. You have doubled your exposure.

The test I use is simple. Run the candidate service through the same worst-case AI scenario for your core solution, and ask whether demand for it survives the fire that would kill the core. If it survives, it is a hedge. If it burns in the same fire, it is more of the same bet wearing a different label.

Paid media against an SEO core is a useful case to think through. You might worry that if AI eats into organic search, it takes paid down with it. I do not think it does, and the reason is the platforms' business model. Google and Meta monetize attention. Even if AI overviews come to dominate the results page and the old blue links recede, the ad business follows the eyeballs onto whatever the new surface turns out to be. The channel changes shape. The advertising does not go away. By the survive-the-fire test, paid is uncorrelated with the specific risk that threatens organic, which is what makes it a real hedge rather than more exposure.

There is a sharper version of this worth internalizing. The most durable hedge is not a different channel. It is the part of the work that AI and the platforms cannot easily capture: strategy, measurement, incrementality, the judgment that sits above execution – staying on the AI operations frontier. A channel can survive while the agency's role inside it gets automated away. Hedge toward the value that is hard to commoditize, and you are protected no matter which channel wins.

Start early, start small, and protect your anchor accounts

If you are going to do this, the timing matters more than founders expect. The instinct is to wait until you are bigger and more stable. The better move is often the opposite. Launch the adjacent line while the revenue at stake is still small, so a stumble cannot take the core down with it.

Consider an SEO firm doing around $6M a year, with a handful of clients on $250K engagements. That same ICP client might carry a paid media budget of $3M.  The temptation is obvious: stand up a paid practice and go straight for that $3M in spend with the existing client. Depending on the fee table, that can yield a similar $250K in fees the SEO firm is familiar with. But those two fees aren’t the same. 

A nascent capability fumbling a program that size doesn't just cost you the new paid opportunity. It puts the stable $250K relationship that is funding you at risk. Also, a major mistake can also require refunding some ad spend (remember the $3MM?) in a way that never enters the convo for SEO services. The danger is not the new revenue you might lose. It is the contamination of the anchor account.

That leads to a sequencing rule that pairs with the "along your customer" idea. Stay inside your ICP for positioning, but do not cut your teeth on your most important relationship. Learn the new capability on a smaller account or a fresh logo inside the same niche, prove it works, and only then graduate it up to the accounts you cannot afford to lose.

How you actually stand the thing up, whether you build with employees, staff with contractors, acquire a firm already doing it, white label through a partner, or run some hybrid, is a separate question with its own risk profile. That is the subject of the next piece, along with the financial guardrails that keep a new venture from becoming an open-ended drain.

The bottom line

Launching a new capability is usually sold as an offensive move, a way to grow the top line. Done with discipline, it is also a defense. It is a way to place measured bets on how AI will reshape your clients' industries and your own, so that you are not a single wager on one future.

The discipline is the whole game. You hedge only if you are actually exposed. You move only once your core is sturdy enough to take the hit. You expand along your customer rather than away from them. You confirm the new line survives the same fire that threatens the old one. And you start small enough that learning is cheap.

Get those right and a second service line stops being a gamble. It becomes optionality, the right rather than the obligation to watch the future unfold and choose your next move. You can double down if it shows promise, throttle back if it does not, or wind it down without taking the core with it. How to build in those escape hatches, the KPIs, the tolerance thresholds, and the art of failing gracefully, is where this goes next.