You're Buying a Bigger Pump for a Bucket With a Hole in It

The Take: Cutting retention budget to fund the acquisition rush is not a strategy. It is a capital allocation mistake dressed up as an AI initiative, and most companies making that trade have never priced what the cut protects.

I have sat in the board meeting where a bigger new-customer number gets waved through in ten minutes and the loyalty line quietly gets trimmed in the same breath. Nobody in the room argues about it. The new-logo figure is exciting. The keeping-people figure is boring. So the boring one gets trimmed and everyone goes back to their spreadsheets feeling like they made a call.

They did make a call. They chose to bleed out faster.

Media buyers are shifting spend into acquisition because AI-driven channels look measurable and fast, while loyalty work reads as a slow, boring line with no obvious near-term payoff. That instinct is now visible in the global data, dressed up as a technology decision rather than what it actually is. Gartner's June 2026 marketing survey found awareness and conversion now account for 62.6% of total media investment, up more than 10% since 2024, while loyalty work has been slashed 29% over the same window to under 15% of the total.

That survey covers North America, the UK and Europe, not Australia. The cost pressure driving it doesn't stop at the border though, and neither does the herd behaviour. When the pack zigs, the instinct is to zig harder.

It's a budget excuse wearing an AI costume. Gartner's 2026 CMO Spend Survey found CMOs are now committing 15.3% of budgets to AI. Only 30% report mature readiness, and 70% admit their internal processes aren't developed enough to scale it. Labour's share of the total jumped from 21.9% to 24.5% in a year. Funds are shifting toward AI and headcount at the same time the funnel machine gets cranked harder, and loyalty work is the item absorbing both moves.

That's an allocation call dressed up as a technology call. The AI line gives everyone permission to stop scrutinising where the funds actually land.

Because almost no one has priced what it delivers in the first place, and you can't defend a figure you've never put on the board. That's the deeper fault sitting underneath this whole shift. Across the 721 Australian-market companies we have scored as at July 2026, the dimension that measures how well a company actually tracks and proves what its marketing is doing sits lowest of the six we assess, averaging 58.5 out of 100. Nearly 44% of scored companies carry that measurement dimension as their single weakest score. When a company can't prove what loyalty work returns, it can't fight for the funding when someone reaches for the scissors. The trim isn't a considered choice. It's the path of least resistance.

I've watched this play out from both chairs. On the agency side, loyalty work was unglamorous, so it got under-sold and under-resourced. On the client side, it was the line with no champion in the room, so it got trimmed first. Same result from opposite directions, and both start with a company that never priced what it was giving up.

Start by pricing what each side delivers. New-customer cost is loud and easy to see. Loyalty value is quiet and easy to overlook. If you can't put a figure on what a returning customer is worth over a year, get that figure on the board before you touch the budget.

Then run the binary test. Take a slice of your loyalty spend and imagine it gone. If revenue holds without it, fine, you've learned something. If revenue falls, you've just found a load-bearing wall you were about to remove. Most companies never run this check because the loyalty line is too quiet to defend itself.

Hold a little in reserve on the AI spend rather than committing the whole line. You don't need 15% of the budget on something 70% of your peers admit they can't run yet. Spend enough to learn and put the rest where the return is already visible.

Resist the short-term flattery. Pulling loyalty budget to inflate this quarter's new-logo count produces a figure that looks good on the statement and a company that's quietly thinning underneath. Cheap now, costly later.

The whole industry is sprinting to fill the top of the funnel faster while the line that keeps existing clients gets thinner. Costs are climbing, AI is crowding every channel and the pack is trimming the one budget item that compounds.

That's the opening. While the herd chases pricier acquisition, you protect what you've already paid for. Treat loyalty as the engine and acquisition as the top-up, not the other way round. A kept client is the cheapest growth you'll ever buy, and right now it's the one line the whole industry is choosing to walk away from. See where your business sits on retention against your industry.

Methodology: The measurement figure above (58.5 average, 44% weakest-dimension share) is drawn from New Rebellion's benchmark dataset of 721 scored Australian businesses across dozens of industries as at July 2026. Each business is scored 0-100 on six weighted dimensions using site, analytics and public signal data. Full methodology at how-we-score.

Is trimming loyalty spend to fund AI ever the right call?

Rarely, and only once you've already priced what loyalty work is worth. Most companies cut the line without knowing what it was protecting.

How much should a business spend on loyalty versus acquisition?

There's no universal ratio. What matters is knowing what each dollar in each bucket actually returns before you move it, not hitting a fixed split.

What's the fastest way to test if loyalty spend is working?

Switch off a slice of it for a defined period and watch revenue. If nothing moves, you've found spend that wasn't earning its keep. If revenue drops, you've found a wall you were about to remove.

Does this data apply to small Australian businesses or just enterprise?

The Gartner surveys cover larger organisations in North America, the UK and Europe. The New Rebellion benchmark data spans small to enterprise Australian businesses, and the measurement gap shows up at every size.