Other Marketing Channels
TikTok, SEO, Affiliates, Podcasts
- Go deeper before wider - taking one channel from 10% to 30% beats going 0% to 5% on a new one
- Before adding any channel, all five framework questions must be yes, including an 8-12 week budget commitment
- TikTok is a creative-volume game: sustain 5-10 new creatives a week or the channel stalls
- Build high-intent blog pages for 'best X for Y' searches - zero cost per click, compounding value
On this page
More channels sound like more opportunity. But the addressable audience within a single well-performing channel is usually so large that you're not even touching the sides. Getting from 10% to 30% of a channel's potential is almost always higher ROI than going from 0% to 5% on a new channel. The default should be depth first. Diversify from a position of strength, not boredom.
This section is about when additional channels actually deserve attention, how to test them without losing focus, and where they can complement the core engine instead of distracting from it.
When to Diversify
If 80%+ of your revenue depends on Meta and Google, you're one algorithm change away from a very bad quarter.
Every mature DTC brand needs channel diversification - diminishing returns at scale, platform risk (iOS 14.5 proved this), and different channels reaching different audiences. But diversification too early kills focus. Dominate two channels, then expand.
Find something that works on one platform, extract the learnings, test whether they transfer. If one channel is delivering exceptional returns, ride it. The time to diversify is when the primary channel is optimised, not when you're bored of running the same playbook.
At Quad Lock, we tested outdoor advertising at scale for the Drive category with billboards, on highways, and in cities. It was run properly: certain states as test markets, others held back as a control group to measure uplift on digital conversions.
The result? Nothing measurable. No uplift on conversions, no bump in the category MER. The only feedback was friends and family saying "hey, I saw you on a billboard." That's an expensive way to impress your mates.
The learning: test new channels with holdout groups and clear success criteria. And be honest when something doesn't work, no matter how good it looks in a pitch deck.
The hooks, ad structures, and testing frameworks in Section 14: Meta Ads - Running & Optimising apply to every channel in this section, not just Meta. The production engine that feeds them lives in Section 17: Content & Creative. Read those first if you haven't already.
The Channel Diversification Decision Framework
Before adding any channel, five questions must all be yes. Gut feel gets you started. Numbers tell you whether to keep going.
- Is my current mix hitting diminishing returns?
- Does this channel reach my target customer?
- Can I commit minimum viable budget for 8-12 weeks?
- Do I have the creative assets this channel requires?
- Can I measure it?
Diversification is great, but make sure you're testing, measuring, and learning. Only go deeper into a channel because the data says it's working, not because someone told you it's the next big thing. The shiny new platform is almost never the answer. The underoptimised existing channel almost always is.
Some of the best "channels" we used at Quad Lock weren't advertising channels at all. We did early licensing and integration deals with platforms like Strava, which put us directly in front of active users at the exact moment they were engaging with their activity. And when we got our events programme properly scaled, that became a genuine acquisition channel too, not a brand awareness exercise. See Section 19: IRL Brand Building for how we built that.
The platforms change constantly, but the principle doesn't: go where your customer's attention already is. Sometimes that's a paid ad, sometimes a licensing deal, sometimes showing up in person. Think beyond the standard advertising channels.
TikTok
TikTok Ads
When TikTok Makes Sense- Target customer typically under 40
- Visually interesting or demonstrable product
- Can produce high-volume UGC-style creative
- Already spending $30K+/month on Meta
- Product works for impulse/discovery
When to Hold Off- Audience skews older
- Product needs education, not entertainment
- Can't sustain 5-10 new creatives per week
- Still scaling Meta and Google
- US regulatory risk exceeds your appetite
Creative approach: Native-looking content wins. Spark Ads (boosted creator content) often outperform traditional creative. Hook in 1-2 seconds. 9:16 vertical, 15-30 seconds.
TikTok is a creative-volume problem, not a targeting problem. On Meta you can still win by out-thinking the auction. On TikTok the algorithm finds the audience for you - your job is to feed it enough native-looking creative that something catches. The brands that struggle here are almost never targeting wrong. They've just run the same three videos into the ground. Weight your library heavily towards UGC and creator content, not polished studio spots (creator sourcing and platform strategy live in Section 18: Social Media), and refresh it constantly: every 7-10 days minimum, and every 3-4 days once you're past $1k/day. A sensible default mix is 80% Spark Ads on creator and UGC content, 20% retargeting to clean up the warm audience.
Spark Ads (boosted creator/organic posts)- Run from the creator's handle with their existing engagement attached
- Roughly 2.4x higher CTR than studio-built ads
- Around 44% better conversion
- Carry social proof the algorithm already rewarded organically
- Native to the feed - they don't read as ads
Studio Ads (brand-built from scratch)- Polished, branded, and obviously an ad
- Lower CTR and weaker conversion
- No borrowed credibility or pre-earned engagement
- Useful as a baseline, but rarely your scalers
- Burn out faster because they never felt native
Typical ranges (at time of writing): CPM $6-$13 (select categories lower), CTR 0.5-1.5%, CPC $0.50-$2.00, ROAS 2-4x. These are evolving as the platform matures and competition increases. Note that 2x ROAS only breaks even at 50%+ contribution margin - calculate your own floor before testing (see Section 14: Meta Ads for the formula).
TikTok Shop & Social Commerce
TikTok Shop has moved from experiment to legitimate commerce channel. In-app checkout eliminates redirect friction (2-5x higher conversion than sending traffic to an external site). Creator-driven selling through affiliate programmes powers discovery. And live shopping, while still early in Western markets, is growing.
But the economics deserve scrutiny.
How it works:
- In-app checkout: Customers buy without leaving TikTok. You connect your Shopify catalogue and manage fulfilment.
- Affiliate programme: Creators promote your products for commission (typically 10-20%). You set the rate, they make the content.
- TikTok Shop Ads: Spark Ads boost organic and creator content. Video Shopping Ads drive directly to in-app purchase.
- Live shopping: Real-time selling through livestreams. Works well for beauty, fashion, and demonstrable products. Requires dedicated team and production.
- Fulfilment: Ship-by-seller (you handle) or Fulfilled by TikTok (TikTok handles). Evaluate based on control, speed, and margin impact.
TikTok Shop Creator Cold-Start Sequence
The affiliate programme is the engine, but you can't switch it on and expect creators to find you. You have to seed it deliberately, and you have to do it in the right order. Don't jump to paid partnerships before you've proven which videos convert - you'll be paying for guesses. Earn the converting videos first with commission-only seeding, then put budget behind them.
Expect roughly 30-50% of the creators you seed to actually post - so seed two to three times more than the coverage you want. And when the posts land, the distribution is brutally uneven: the top 20% of creators typically drive around 80% of the GMV. The job isn't to get everyone posting. It's to seed widely enough that your handful of stars surfaces, then concentrate budget and paid partnerships on them.
Commission rates sit in the 10-20% band (the US average is around 13%). That comes off the top before your other platform fees and COGS, so model it into the true cost-per-sale picture in the economics callout below - the take-home maths is unforgiving if you stack a generous commission on a thin-margin SKU.
GMV Max (TikTok Shop Ads Automation)
GMV Max is now the default way to scale TikTok Shop ads. It has largely replaced manually building and managing individual Shop ad campaigns. You set a SKU, a budget, and a ROAS target, and TikTok's system handles the rest: it pulls from your organic posts, creator videos, and product feed, then automatically tests and serves whatever combination converts. Less hands-on-the-dials, more feeding the machine the right inputs.
The inputs are where you win or lose. GMV Max is only as good as the creative pool you give it, so feed it your best-performing creator content - the videos that already earned watch time and sales organically. This is not a channel you bootstrap from a standing start.
GMV Max optimises against the creative you give it. If you switch it on with no library of creator content that has already worked organically, you are asking an automated system to scale guesswork. Prove the videos convert through your affiliate and organic posting first (see the TikTok Shop & Social Commerce section above), then hand the winners to GMV Max to scale.
Set your targets realistically and let them tighten. Run a softer ROAS target while the system learns, then step it up as the data builds.
On budget, start around $50-$150/day per active SKU. Enough to give the system room to find winners, not so much that you are buying volume before you know the creative converts. Scale the budget per SKU only after the ROAS holds at your day-30 target.
The economics:
The platform takes a referral fee (6% standard as of mid-2026, with a 3% new-seller promotion for the first 30 days, but rates have moved multiple times). Verify current rates in TikTok Shop's seller policy before modelling. Creator affiliate commissions of 10-20% (US average ~13%) are on top of that. Payment processing fees are typically separate in the UK/EU (around 2.9% + $0.30 per transaction) or blended into the referral fee in the US (effective rate ~1.02% in some cases). Factor in fulfilment and any ad spend, and the total channel cost-per-sale (before product cost) often lands at 25-40% of revenue.
Compare that to your DTC site where you own the customer data, the margin, and the relationship. TikTok Shop is a discovery and acquisition channel, not a margin channel.
Who it works for:
- Impulse-purchase products that are visually demonstrable
- AOV under $50 is the sweet spot (higher price points can work but conversion drops)
- Products where seeing it in use drives the purchase decision
- Brands that can produce or source high-volume short-form video content
Who it doesn't work for:
- Complex products requiring education before purchase
- High-AOV products where customers need time to decide
- Brands that can't sustain video content (TikTok is a content-first platform - no content means no sales)
- Categories where brand trust matters more than discovery (premium skincare, health devices)
The DTC vs TikTok Shop Tension
The strategic tension is real and worth being honest about:
- Customer ownership: TikTok owns the customer data. You get the sale but limited access to the relationship. You can't email them, you can't retarget them, you can't build LTV the way you do on DTC.
- Margin erosion: Platform commission + affiliate commission can eat 20-30% of revenue. On a 60% contribution margin product, TikTok Shop typically leaves you with roughly 30-40% margin after platform and creator costs, before ad spend - and ad spend is what pulls many sellers under water.
- Brand control: Creator content is authentic but unpredictable. You can brief creators, but you can't control what they say. For some brands this is liberating. For others it's terrifying.
- Cannibalisation: Are TikTok Shop sales genuinely incremental, or are existing customers (who would have bought on your DTC site at full margin) buying through TikTok instead?
The strategic answer: Use TikTok Shop for discovery and acquisition, then move customers to DTC for retention and LTV. Include a card insert in TikTok Shop orders driving to your website for loyalty, exclusive products, or better bundles. Track what percentage of TikTok Shop buyers ever purchase on your DTC site. That number tells you whether TikTok Shop is building your brand or just renting someone else's audience.
TikTok Shop Metrics
| Metric | What to Track | Why |
|---|---|---|
| Gross Merchandise Value (GMV) | Total sales through TikTok Shop | Top-line channel performance |
| nCAC via TikTok | Cost to acquire a genuinely new customer | Incrementality check |
| Creator ROI | Sales per creator vs commission paid | Affiliate programme efficiency |
| DTC crossover rate | % of TikTok Shop buyers who later buy on DTC | Long-term channel value |
| Margin after all fees | Revenue minus COGS, platform fees, affiliate commission, fulfilment | True profitability |
When to Start
The honest take: TikTok Shop is a real channel with real revenue potential, but it comes with real trade-offs. Lower margins, limited customer ownership, and dependency on a platform that could change its rules or availability at any time. Treat it like Amazon - useful for reach and discovery, dangerous as a primary channel. Your DTC site should always be the centre of gravity.
Measurement & Incrementality
Once you're running more than two channels, the measurement problem stops being a footnote and becomes the whole game. The channels in this section - podcast, TikTok Shop, affiliate - are exactly the ones standard attribution gets most wrong, which is why they get killed first in budget reviews. Usually unfairly.
Last-click attribution systematically over-credits branded search and under-credits the channels that create the demand in the first place - podcast, TikTok, top-of-funnel social. Someone hears your host-read ad, searches your brand a week later, clicks the branded search result, and buys. Last-click hands the credit to search. The podcast that actually drove it shows a CAC that looks terrible, you cut it, and your branded search volume quietly falls off a cliff a month later. Treat any platform's self-reported ROAS as directional, not truth.
You don't need all of these on day one, but you need more than last-click. (1) Geo-lift and holdout tests - turn a channel off in some regions, leave it on in others, and measure the real difference. Haus, Measured, Lifesight and Northbeam all run these. (2) Media mix modelling (MMM) - statistical attribution across the whole mix; Google open-sourced Meridian in early 2025 if you'd rather not pay for a vendor. (3) The low-tech one that works for hard-to-track channels: unique promo codes plus a post-purchase 'how did you hear about us?' survey. For podcast, TikTok Shop, and affiliate especially, that survey often tells you more than your analytics dashboard does.
This ties straight back to the diversification framework at the top of the section - 'can I measure it?' was question five for a reason. A channel you can't measure honestly is a channel you'll eventually defund for the wrong reasons. The full attribution and incrementality methodology - geo holdouts, MMM, the measurement stack - is the job of Section 27: Measurement & Data; the unit-economics discipline that sits underneath it lives in Section 26: Finance & Unit Economics.
SEO & Organic Search
Search Engine Optimisation (SEO) takes 6-12 months but compounds. For most brands under $2M, time is generally better spent on paid media, but get the foundations right to avoid costly rebuilds.
Technical Foundations
Do once, maintain quarterly:
| Element | Target |
|---|---|
| Site speed (LCP) | Under 2.5 seconds |
| Mobile-first | Responsive, fast, thumb-friendly |
| Sitemap | Submitted in Search Console |
| Schema markup | Product, review, FAQ where relevant |
| URLs | Clean, descriptive, no parameters |
| Internal linking | Logical hierarchy, no orphan pages |
| Canonical tags | One per page, self-referencing |
| HTTPS | Non-negotiable |
| Hreflang | Set up if selling internationally |
Most DTC product pages are only partly machine-readable - the price, availability, and review data a search engine (and increasingly an AI engine) needs to surface you is either missing or incomplete. Fixing it is dull, one-off work with outsized payoff. Mark up every product page completely: name, offers (price and availability), aggregateRating, and a real review collection. This is the single task that lifts both classic SEO and AI discoverability at once, because both pull from the same structured, extractable data. If your product data isn't machine-readable, you're invisible to the systems that are increasingly deciding what shoppers see.
Content That Works for DTC
Publish 2-4 high-quality articles per month targeting keywords with 200+/month volume and under 40 difficulty. Focus on:
One of the most underrated SEO plays: build detailed blog posts as landing pages for specific use cases. Target long-tail, high-intent keywords where someone has a specific problem and is actively searching for a solution. These pages sit on your site as genuinely useful content, improve your domain's overall quality rating, and bring in ready-to-buy customers day after day. Unlike paid traffic, the cost per click is zero and the value compounds over time.
One of our best SEO plays at Quad Lock was building blog posts as high-intent landing pages for specific use cases. We'd write a post like "Best BMW R1200GS iPhone/Galaxy Mount" targeting the exact long-tail keyword a rider would search. It wasn't a blog post in the traditional sense. It was a product page disguised as helpful content: keyword-rich title, hero image showing the product on that specific bike, social proof from real customers, a product video, and a direct link to buy.
Someone searching "best BMW R1200GS phone mount" has already decided they want one. They just need to know which one. We gave them the answer, backed it with proof, and made it easy to buy. We built dozens of these across different bike models, car models, and use cases. Each one sat on the site bringing in ready-to-buy customers day after day, at zero cost per click.
The High-Intent Blog Post Recipe
The beauty of this approach is scale. Every vehicle model, every use case, every "best X for Y" query is a new page. Each one targets a different long-tail keyword, brings in a different audience, and compounds over time. One page won't move the needle. Fifty will.
As more shoppers start using AI tools and LLMs to research purchases, the content that answers specific questions well is likely to become even more important. Early signs suggest that LLMs tend to draw from the same kind of detailed, authoritative content that ranks well in traditional search. Nobody knows exactly how this will play out, but building a library of genuinely useful, use-case-specific content is a bet that's hard to lose regardless of how discovery evolves.
Answer Engine Optimisation (AEO) for AI Visibility
The callout above hints at it; this is the part you should plan for deliberately. Search is splitting into two jobs - ranking on the page, and getting cited by the AI answer that now sits above the page. AI Overviews and chat-based engines are quietly eating the click. Roughly 60% of searches now end without a click to any website, and when an AI Overview appears it can cut click-through to the top organic results by 30-50% (some 2025 studies put it higher still). If your entire SEO strategy assumes someone scrolls to a blue link, you're optimising for a shrinking surface.
The good news: AEO is not a separate discipline you bolt on. It's the same bet as the high-intent content above, with the structure tightened. Write content an AI can quote cleanly - direct answers up front, specific numbers in the body, claims attributed and sourced. The detailed, authoritative, use-case-specific pages that already rank are the ones engines pull from.
This is the part most SEO advice misses. One of the biggest drivers of whether an LLM mentions your brand isn't your backlink profile - it's how often people search for your brand by name, and how often third parties mention you. The numbers back it: Ahrefs' 2025 study of 75,000 brands found branded web mentions correlate with AI visibility roughly 3x more strongly than backlinks do, and YouTube reviews were the single strongest factor of all. Platform quirks matter too - nearly half of Perplexity's citations come from Reddit. Off-site authority does the rest: most of the sources LLMs cite are third-party - Reddit threads, review sites, press - not your own domain. Which means the highest-leverage AEO work isn't a meta tag. It's the brand-building, PR, and review-generation covered in Section 17: Content & Creative and Section 19: IRL Brand Building. Build the brand, and AI visibility comes with it.
The practical implication: don't treat AEO as a new channel to staff. Treat it as another reason to do the brand and content work you should already be doing - and to get your product data machine-readable (see the markup callout above). E-E-A-T (experience, expertise, authoritativeness, trust) is the currency. Real authorship, real reviews, and a brand people search for by name are what get you quoted.
Your AI-Visibility Checklist (Shopify)
If you're on Shopify, most of the plumbing is already done for you: the platform allows every major AI crawler by default, auto-serves the agent-discovery files, and syndicates eligible products into ChatGPT, Perplexity, Copilot and Google's AI surfaces through Shopify Catalog and Agentic Storefronts - on by default for eligible US-selling merchants since March 2026. Your job isn't integration work. It's a short verification pass, then product-data quality forever after:
- Review your Agentic Storefronts settings (Shopify admin, Sales channels, then Agentic). You're most likely already included - decide deliberately which AI channels get your catalogue and which products to exclude (set them to Unlisted). Note there's no full opt-out of Shopify Catalog itself, only per-channel and per-product controls.
- Complete your policy pages (Settings, then Policies: terms, privacy, returns). This is an eligibility requirement for the AI channels, and assistants read these pages when customers ask about returns or shipping.
- Write product data as answers, not keywords. Titles, descriptions and option attributes get read by a model answering "which X actually works for Y?" - the same shift as the Amazon listing advice in Section 23. Use Shopify's Catalog Mapping to expose metafields the feed would otherwise miss.
- Get review schema onto your product pages. Default themes emit basic product markup but NOT aggregate star ratings - a review app (Judge.me, Yotpo, Loox) closes that gap, and on-page reviews, FAQs and Q&A content measurably separate the products AI assistants put first.
- Install Shopify's free Knowledge Base app and curate how your brand, FAQs and policies read when an AI describes you.
- Enrol in Perplexity's Merchant Program (free, no transaction fees): discovery happens automatically via Shopify Catalog, but enrolling adds priority indexing, in-chat checkout eligibility and search-trend data.
- Enable guest checkout and an express wallet (Google Pay or PayPal) - the observed eligibility requirements for Google's agentic checkout in AI Mode and Gemini.
- Register with Bing Webmaster Tools as well as Google Search Console (two-click import from GSC). ChatGPT's search layer leans on Bing's index; most brands feed Google and forget Bing exists.
One thing NOT to spend money on: llms.txt generator apps. Shopify serves the file natively now, and the log-file studies are blunt - the overwhelming majority of published llms.txt files are never fetched by anyone, and Google has said it doesn't use the standard. The feeds, the crawl access and the third-party mentions above are what move the needle.
On a headless or custom stack, the AI-crawler plumbing Shopify handles for you is YOUR job - and the defaults work against you: Cloudflare has blocked AI crawlers by default on new sites since mid-2025, and other CDN/WAF vendors ship similar settings. Check three things: your CDN's bot-management settings actually allow the AI crawlers (GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot), your robots.txt doesn't disallow them, and your product schema is complete. We ran exactly this audit on this site (not a Shopify store) in July 2026 and found our CDN silently bouncing Perplexity's crawler 60 times a day - real demand, locked out, and invisible on every dashboard we normally looked at.
However you're hosted, measure what comes back: GA4 added a native "AI Assistant" traffic channel in May 2026, and Shopify tags AI-channel orders in admin - the details live in Section 27: Measurement & Data. AI referrals are still small in absolute terms, but they've grown roughly 14x since late 2024 and convert around 40-50% better than average - the classic early-channel profile: too small to staff, too high-quality to ignore, and cheapest to win while your competitors haven't looked.
SEO Tools
| Tool | What It Does | Cost |
|---|---|---|
| Google Search Console | Track rankings, impressions, clicks. Monitor indexing issues. Non-negotiable. | Free |
| Google Trends | Monitor search volume trends over time, spot rising queries, validate new content ideas before you write them | Free |
| Google Analytics (GA4) | Traffic sources, landing page performance, conversion paths | Free |
| Ahrefs | Keyword research, backlink analysis, competitor content gaps | Paid |
| Screaming Frog | Technical crawl audits, broken links, redirect chains | Free up to 500 URLs |
| Surfer SEO | On-page optimisation scoring, content briefs, keyword density | Paid |
| ChatGPT/Claude | Draft content briefs, generate title variations, expand keyword lists | Free/Paid |
Podcast Advertising
Podcast advertising can work for the right brand and audience, but apply the same discipline you would to any channel: is it moving the needle? Unique discount codes, vanity URLs, post-purchase survey attribution. If you can't prove it's contributing within 90 days, move the budget to something that is.
A host reading your ad in their own words drives roughly 31% higher purchase intent than a producer-read or inserted spot, and host endorsements pull two to three times the response of a generic read. The reason is the same one in the founders-note below: you're borrowing the host's credibility with their audience, not just buying airtime. So spend where that credibility is strongest - endemic shows whose listeners are your customers - and read the host into the brand, don't just hand them a script.
Every podcast placement needs its own promo code or vanity URL, full stop - without one you're flying blind, and last-click will badly under-credit the channel (see Measurement & Incrementality above). For efficient host-read campaigns on the right show, DTC CAC can land in the $24-58 range, but treat that as a best case, not a default - broad podcast CPA more commonly sits well above it. When you negotiate, push for terms tied to incrementality and performance, not raw impressions. Impressions are what the show sells. Acquired customers are what you're buying.
We saw this dynamic repeatedly at Quad Lock with media partners and personalities in our categories. The ad read itself could work, but the bigger upside came when the person already had credibility with riders or cyclists and we extended the relationship beyond a single placement. We'd use that trust across the website, paid creative, and broader brand activity. The host stopped being just inventory and became part of the brand story. That's where the compounding value lived.
YouTube Creators
The same credibility logic applies to YouTube creators, with one big difference: the content compounds. A long-form review or "best X for Y" video from a creator your customers already trust keeps surfacing in search and recommendations for years, doing the work of an always-on salesperson. Map the top creators in your categories, get product into their hands, and build the relationship before you buy the integration - a genuine review outperforms a scripted ad read here just as it does on podcasts. Put affiliate links or unique codes behind every placement (see below) so the channel earns its budget on measured results, not view counts.
Affiliate & Partnership Programmes
An affiliate programme is one of the few channels that only costs you money when it works - you pay on the sale, not the click. That makes it tempting to treat as free money. It isn't. A programme left to run on autopilot fills up with coupon and deal sites that intercept customers who were going to buy anyway, and you end up paying commission on conversions you already owned. Run it deliberately or don't run it.
Commission benchmarks. The standard sits at 10-15% of order value, or a flat $10-15 bounty per new customer if you'd rather pay for incrementality than volume. By category, apparel tends to run 8-15% and beauty 10-18%. Deal and coupon sites should be capped hard - 0.5-1.5% - because their job is to skim, not to acquire.
| Partner type | Typical commission | What you're actually buying |
|---|---|---|
| Content / review sites | 10-15% | New audiences, durable SEO-backed pages |
| Influencer-affiliates | 10-15% (or flat bounty) | Awareness plus trackable conversion |
| Loyalty / cashback | 2-5% | Incremental checkout nudge |
| Deal / coupon sites | 0.5-1.5% | Mostly interception - weight low |
Platforms. Impact.com, ShareASale, CJ, and Rakuten are the established networks for managing partners, tracking, and payouts. If a meaningful share of your revenue runs through Amazon, Levanta is the affiliate platform built for that.
The brands getting real leverage here don't run influencer and affiliate as separate budgets. They stack them - the creator drives awareness with their content, and the affiliate link captures and attributes the conversion. The creator content does the persuading; the affiliate infrastructure does the measuring and paying. This is the same whitelisting logic from the paid channels - prove it with content, then put trackable structure behind the winners.
What to expect. A new programme typically returns around 4:1 to 6:1 ROAS in year one as you recruit partners and weed out the freeloaders. As it matures, the right move is to tighten efficiency targets, not loosen them - prune low-incrementality partners and concentrate spend on the content and influencer-affiliate partners genuinely bringing new customers. The whole game is the incrementality question the checklist already flags: are these new customers, or conversions you'd have won anyway?
Referral programmes. The customer-powered version of the same pay-on-results idea: a two-sided incentive ("give $20, get $20") that rewards an existing customer for bringing a friend and gives the friend a reason to try you. Referred customers arrive pre-sold by someone they trust, so they typically convert better and retain better than paid traffic. Keep the mechanics simple - one clear reward each side, easy sharing, automatic tracking through tools like ReferralCandy, Friendbuy, or your loyalty platform - and treat it as a retention asset as much as an acquisition channel (Section 21: Customer Retention & Loyalty).
Section 16 Checklist
Go from reading to doing.
You've read the section. Sign in to score your ecommerce brand with the Health Check, track your progress across 472 checklist items, and get your tools and history in one dashboard. Free, and always will be.