How to attract customers and grow a business on social media

Executive overview

Viral views don't pay the bills — a video with 15 million views can earn less than one with 30,000 if it targets the wrong audience. The real goal is building a niche audience of buyers, then owning that relationship through email before the algorithm changes again.

High-view content and high-converting content are almost never the same video.

Platform trade-offs

• YouTube: evergreen search traffic, passive ad revenue, strongest audience connection — but requires capital and storytelling skill
• Instagram Reels: fastest organic growth right now; daily posting compounds quickly; Stories + DMs are the best direct sales tool
• TikTok: viral reach with weak monetisation; algorithm ignores follower count; hard to convert unless you need only 1–2 clients/month

YouTube growth strategy

• Define your primary goal first: celebrity or revenue — they require different approaches
• Research the top 5 channels in your niche; focus on their best-performing videos from the past 12 months only
• Use search-based topics early; 80% of eventual traffic comes from recommendations, but search anchors a new channel
• Content mix: 40% proven formats in your niche, 40% expert Q&A on questions you already get asked, 20% experiments or direct sales
• Use Perplexity (not ChatGPT) to generate 100 topic ideas, then filter by: million-view potential, trust-building, clarity of problem, follow-up opportunity
• Channel name can itself be a keyword — a channel named "Silicon Valley Girl" landed brand deals at 20k subscribers because advertisers wanted that audience

Video formats and production

• Talking head, reaction, solo podcast, interview, documentary, news, review, tutorial — match format to what's already working in your niche
• Phone + good mic + good lighting beats expensive gear; DJI Pocket Pro recommended for portable shooting
• Edit with iMovie, Adobe Premiere, or CapCut; outsource editing for $25/hr or less
• Thumbnails: photo + phrase; $75–$100/thumbnail if outsourcing
• Research tools: VidIQ, TubeBuddy, SocialBlade

Niche discipline

• Pick 10 topics before you launch — this defines your content strategy for the first few months
• 60%+ of content must stay within your niche; adding unrelated topics confuses the algorithm and repels subscribers
• Teach to a "fifth-grade level" in your topic — oversimplification drives completions; over-complexity kills watch time
• Views ≠ money: high-converting niches include test prep, tax strategy, real estate, job hunting — not lifestyle or entertainment

Instagram-specific tactics

• Post every single day to trigger current algorithm prioritisation
• First frame / first phrase determines whether anyone watches — this matters more than on YouTube
• Like content from accounts in your niche; the algorithm uses this to connect you to their audience
• Use trending audio formats to accelerate reach
• DM automation via ManyChat: end a Reel with a keyword CTA ("DM me 'video'"), collect the email automatically, deliver the lead magnet

Converting viewers into customers

• Email list is the only audience you own — start collecting before you have a product
• ConvertKit and Beehive both offer free tiers up to 2,500 subscribers; each email generates ~$1,000+ per send (at scale)
• Product ladder: free lead magnet → low-ticket trust builder → mid-ticket course/seminar → high-ticket 1-on-1
• Sell on YouTube via pinned links; sell on Instagram via Stories and link-in-bio (use Beacons AI)
• Once you understand your customer profile, paid ads (Facebook/Instagram) become the highest-revenue channel — social builds connection, ads close the sale
• Create lookalike audiences from your email list for paid acquisition

Consistency and mindset

• YouTube is a 12-month commitment: 4 videos/month minimum to build a traffic-driving channel
• Shorts: post 2–3/month regardless of strategy — the platform may shift toward short-form; learn the format now
• Track CTR (target 7–8% on recommendations) and audience retention drop-off points
• Reputation is permanent — assess risk before experimenting, test with small groups first, don't feed trolls