10 growth tactics that never work, and three frameworks that do

Executive overview

Growth teams are failing at high rates — not because growth is hard, but because the field is flooded with out-of-context hacks borrowed from one company's specific situation. The recurring culprits are predictable patterns: hiring too early, redesigning instead of optimising, copying competitors, and treating every decision as an experiment.

The core insight: growth amplifies strong product-market fit — it cannot substitute for it, and no tactic, team, or redesign will fix a fundamentally broken growth engine.

The 10 tactics that never work

1. Hiring a growth team too early. Founders must drive growth to the first $1–10M ARR. Growth teams need volume (data) and retention (PMF) to function. Without both, they can't experiment meaningfully.
2. Hiring a growth team to reverse a decline. Growth can lift results by 10–15% at best. If the business is shrinking, the root cause is product or go-to-market — not a missing growth team. Stabilise the decline first.
3. Rebranding or redesigning the homepage to drive acquisition. Every rebrand is a performance step backward. Optimisation to recover takes 3–6 months minimum. Never promise acquisition lift from a redesign at launch.
4. Copying competitor flows. You don't know if what you're seeing is their real experience, a test variant, or a legacy page nobody maintains. Copying produces mediocrity and skips the design and research that makes an experience actually work.
5. Assuming your problem is unique. Almost no growth problem is new. Find people who've solved it — the cost is time on LinkedIn or a paid advisor session. Re-engineering from scratch wastes cycles the market won't forgive.
6. Prioritising rented channels (SEO, SEM, paid social) over earned ones. Rented channels depend on algorithms and platforms you don't control. Earned channels — virality, referrals, user-generated content — can't be competed away. Dropbox gets 50%+ of acquisition through sharing alone.
7. Locking into one growth model. Every growth loop has a lifespan (typically 5–7 years). Overlay new models — product-led, sales-led, marketing-led — before the current one saturates. Allocate 20–25% of growth team time annually to building the next loop.
8. Not hiring advisors. Advisors compress failure cycles. Before hiring, run a paid workshop on a real problem and evaluate whether they added value. Reassess monthly. Some advisors are useful for three months; others for years.
9. Testing everything. If a test needs more than a month to reach sample size, don't run it — use pre/post instead. Experimentation paralysis kills velocity. Trust intuition on low-traffic or low-stakes changes; reserve rigorous tests for high-traffic, high-revenue real estate.
10. Lightning round — four more common mistakes:
    • Color optimisation: shade of blue doesn't move the needle in 2025. Pick one, move on.
    • Third-party auth: adds marginal conversion lift but rarely creates incrementality in activation or retention. Do it for UX, not growth metrics.
    • One-off emails: a single email at 25–40% open rate will never produce meaningful lift. Think in sequences and communication strategies, not single sends.
    • "Simplify onboarding" as a roadmap item: simplification is a solution, not a problem. Define the actual problem (confusion, drop-off, lack of education) first. Removing steps without a clear problem statement produces nothing.

Three frameworks worth knowing

• Growth loops (Reforge / Brian Balfour, Casey Winters, Andrew Chen). Think in action → reaction → action cycles, not funnels. Loops compound; funnels don't.
• Race car framework (Lenny Rachitsky and Dan). Separates growth work into engines (loops), fuel (paid budget), turbo boosts (one-time events), and lubrication (ongoing optimisation). Clarifies what type of initiative you're actually building.
• Adjacent user theory (Bangaly Kaba). Identifies users just outside your ICP whose experience can be optimised without expanding product-market fit — a growth team responsibility distinct from core product work.

On earned vs. rented channels

• Rented: organic search, paid search, paid social. Fast to start, always competing in someone else's arena, subject to algorithm changes.
• Earned: virality, referrals, user-generated content, community. Expensive in people and product resources to build, but competitors cannot replicate them once established.
• With AI interfaces replacing search as the primary discovery UI, rented channel ROI is set to deteriorate further. Teams not investing in earned channels now will face rising acquisition costs with no alternative.

On career optionality

• Full-time employment is one monetization model, not the only one. Others include advising, consulting, fractional roles, interim engagements, courses, newsletters.
• Optimise for career optionality — the ability to choose what fits your skills, life stage, and interests — rather than for a specific title.
• Earning optionality still requires depth built through real operator experience. But once that foundation exists, defaulting exclusively to full-time roles is a constraint, not a strategy.
• Evaluate each opportunity by asking: does this increase my option pool, keep it flat, or shrink it?