Building the Business Case for AEO: Budget Framework, ROI Model and Board Slides

What does an AEO business case need to prove?

An AEO business case has to prove three things: that a meaningful share of your category's buying research now happens inside AI assistants, that your brand's presence in those answers can be moved deliberately rather than by luck, and that the movement connects to pipeline at a cost per opportunity competitive with your existing channels. Everything else in the deck is supporting evidence for those three claims.

Most rejected proposals fail on the second point. Marketing leaders arrive with convincing evidence that AI search is growing and no evidence that their own team can change their position within it. A finance partner reads that as a bet on a trend rather than an investment in a capability, and prices it accordingly, which usually means a pilot budget too small to produce a signal.

The fix is to run a small baseline before you ask for anything. Two to four weeks of prompt tracking across 100 to 200 category questions gives you a starting citation share, a competitive gap and a list of prompts where you should obviously appear and do not. That gap, expressed as a number, is the strongest single slide in the deck.

How much should an enterprise budget for AEO?

Enterprise AEO budgets typically sit at 8 to 15 percent of total search spend in the first year, which for most mid-to-large B2B organizations means 150,000 to 500,000 dollars annually. That range covers prompt monitoring and tooling, content restructuring, entity and schema work, digital public relations for third-party corroboration, and the analyst measurement time needed to report credibly.

Allocate roughly 40 percent to content, 25 percent to off-site corroboration, 20 percent to technical and entity work, and 15 percent to tooling and measurement. Tooling is the smallest line and the one most often overfunded. Monitoring platforms in this space typically cost 15,000 to 60,000 dollars annually at enterprise scale, and buying the platform without funding the content and PR work behind it produces a dashboard that measures a problem nobody is resourced to fix.

Phase the ask. A first-year request of 8 to 10 percent of search spend, with an explicit trigger to step up to 15 to 20 percent on hitting a citation-share threshold, converts far more often than a single large number. Finance teams approve staged commitments with defined exit points more readily than they approve conviction.

The ROI model: what a single AI citation is actually worth

Value a citation the same way you value an organic ranking, through a chain of estimates rather than a single figure: tracked prompt volume, multiplied by citation share, multiplied by the rate at which a cited answer produces a visit or an unattributed brand consideration, multiplied by your opportunity conversion rate and average deal value. Build the chain in a spreadsheet the CFO can edit, because a model they can stress-test is far more persuasive than a model they must accept.

Use conservative inputs. Assume only 10 to 20 percent of assistant answers produce a measurable session, assume assistant-referred sessions convert at two to three times generic organic rather than the higher multiples you may observe early, and hold average deal value flat. In most enterprise models with a 40,000 to 150,000 dollar average deal size, moving citation share from 10 to 30 percent on a 250-prompt set produces somewhere between 8 and 25 incremental opportunities per quarter.

Include a brand-influence term but label it clearly as unattributed. A defensible way to present it is to report the share of closed-won opportunities where an assistant appears anywhere in the recorded touch path, then note explicitly that this is directional. Calling out the soft number as soft buys credibility for the hard numbers next to it.

Present a downside case alongside the base case. A model that shows only the optimistic path invites finance to discount every input, while a model with an explicit floor scenario, in which citation share moves half as far and assistant referrals convert at ordinary organic rates, tends to be accepted at face value. In most enterprise proposals the floor case still clears the internal hurdle rate, and demonstrating that is more persuasive than any headline return figure.

Payback on AEO typically lands between 7 and 12 months

Payback in enterprise AEO programs typically arrives 7 to 12 months after the first content and entity changes ship, which is faster than most classic SEO programs and slower than paid media. Citation movement begins around 60 to 90 days, pipeline attribution becomes readable around month five or six, and closed revenue lands one sales cycle later.

Sales cycle length drives the variance more than program quality does. An organization with a 60-day cycle can show closed revenue inside three quarters, while a 9-month enterprise cycle will show influenced pipeline in year one and closed revenue in year two. Set that expectation explicitly in the approval meeting, because a business case that promises revenue on a timeline the sales cycle cannot support will be judged a failure regardless of how well the program performs.

Cost per opportunity is the fairer year-one metric. Most programs we see settle at 40 to 70 percent of the blended cost per opportunity from paid search once the channel matures, largely because assistant-referred visitors arrive later in their research with a narrower vendor set already in mind.

Three-Horizon Funding Case: a framework for staged AEO investment

We structure enterprise proposals as a Three-Horizon Funding Case, which maps the investment to what each horizon can honestly prove. Horizon one covers months zero to six and buys proof: baseline measurement, restructuring of the 20 to 40 pages that already attract category demand, entity cleanup, and a repeatable monthly citation report. The only commitment made here is movement in citation share, not revenue.

Horizon two runs months six to eighteen and buys scale. This is where the corroboration budget grows, where content coverage extends across segments and regions, and where the CRM connection turns citations into reportable influenced pipeline. Roughly 60 percent of total program spend belongs in this horizon, and it is the horizon most business cases underfund because the early wins came cheaply.

Horizon three spans months eighteen to thirty-six and buys durability: proprietary data, original research, executive thought leadership and analyst relationships that competitors cannot replicate by rewriting a landing page. Presenting the case in three horizons lets a CFO approve horizon one without approving all of it, which is usually the difference between a funded program and a deferred one.

Which five slides does the board actually need?

Five slides carry the decision. The first shows the behavior shift with your own baseline data: the share of your tracked category prompts already answered without a click, and where your brand sits in them. The second shows the competitive gap, naming the three or four rivals cited more often than you and the specific prompt clusters where they win.

Slide three is the model, presented as a single chain from prompts to opportunities with editable assumptions and a visible downside case. Slide four is the phased budget mapped to the three horizons, with the step-up trigger stated as a number rather than as a milestone description. Slide five is the operating plan: who owns the program, which teams contribute, and what gets reported monthly.

Keep the deck under ten minutes of speaking time and put the technical detail in an appendix. Boards approve clear ownership and defined exit conditions. They rarely approve schema strategy, and a proposal that spends its airtime on retrieval mechanics tends to leave the room without a decision.

Rehearse the two objections that usually arrive before slide five. The first is why this cannot wait a year, answered with the observation that corroboration assets take two to three quarters to build and cannot be purchased quickly later. The second is why the incumbent agency cannot simply absorb the work, answered with the specific capability gap in measurement and entity resolution rather than with any criticism of the current partner.

Answering the CFO's three hardest questions about AEO

Expect three questions, and prepare exact answers. The first is incrementality: are these opportunities genuinely new, or would the buyer have found you anyway. Answer with a holdout comparison, tracking two matched prompt clusters where you invest in one and hold the other flat for a quarter, then report the difference in citation share and referred sessions.

The second is cannibalization: does this simply move traffic from organic to assistants. The honest answer is that some substitution occurs, and the useful response is to report total qualified sessions and opportunities across both surfaces rather than defending organic volume in isolation. Programs that measure the combined number usually show flat sessions and higher conversion, which is the correct story to tell.

The third is platform risk: what happens when the assistants change their retrieval behavior. The defensible position is that entity clarity, content quality and third-party corroboration are the durable inputs across every generation of these systems, so the assets you build survive the interface changes. This is also the natural place to note that a partner such as Lemniscate Growth structures these programs around pipeline outcomes, so the reporting line holds even when the underlying platforms shift.

Why approved AEO budgets get cut in the second year

Approved budgets get cut when the program reports visibility instead of pipeline. A citation-share chart with no CRM connection reads as a vanity metric during a tight planning cycle, and it competes badly against channels reporting cost per opportunity. Wire the CRM connection in the first six months, even imperfectly, because a rough attribution model beats an elegant visibility dashboard in a budget defense.

A second failure is the missing baseline. Teams that never captured a starting position cannot demonstrate movement, only a current state, and a current state is indistinguishable from luck. Capturing the baseline costs two to four weeks and protects the entire investment, which makes it the highest-return activity in the whole program.

The third is ownership drift. When the accountable owner changes roles and the program becomes everyone's contribution, monthly reporting slips, then the forum stops meeting, then the line item is reallocated. Naming a successor in the original proposal sounds procedural, but in enterprise engagements it is one of the clearest predictors of whether an AEO program survives its second planning cycle.