How concentrated is your marketing
Concentration is easy to sense and easy to misjudge. Measure it from the business's own records, over a period long enough to smooth out seasons, and look at three things.
- New customers by source. Which channel or platform did each new customer come from? Use the CRM or order records, the analytics source, and the answers to a "how did you hear about us?" question where tracking is weak.
- Revenue by source. First orders and repeat orders separately. A business may win most first orders through ads while most revenue comes from repeat customers it already owns, which is a very different risk.
- Enquiries or leads by source, for businesses that sell through conversations rather than checkouts.
Then split each view two ways: by channel (paid search, paid social, organic search, marketplace, email, referral) and by platform (the company that controls the rules). The second view often reveals more. Organic and paid search on the same search engine are two channels but one platform; a marketplace listing and the marketplace's own advertising are two channels that both depend on one marketplace's rules.
| View | What to count | What it reveals |
|---|---|---|
| New customers by channel | First orders or new accounts, by channel | Where growth comes from |
| New customers by platform | The same, grouped by the company that controls each channel | How many sets of rules the business depends on |
| Revenue by source | First and repeat revenue, separately | Whether owned relationships already cushion the business |
| Enquiries by source | Leads or enquiries, by channel and platform | Concentration in businesses that sell through sales conversations |
| Cost per customer by source, over time | CACCost to acquire one new customer. in the main channel, month by month | Whether the main channel is getting more expensive |
There is no share above which a business is too concentrated. The test is the consequence: what would happen to cash, staff and stock if the main source halved for a month, or stopped? A business with savings, a strong repeat base and a second route can absorb a shock that would close a business without them.
What concentration risks
A channel that works can change for reasons that have nothing to do with the business. The main risks fall into five groups.
Restrictions and policy changes. Every platform enforces rules, and enforcement can stop a channel abruptly. The site's guide to Google Ads policies and disapprovals explains that repeated or serious breaches can suspend an account, that suspensions are far harder to reverse than disapprovals are to fix, and that a change to the website can disapprove ads nobody in the account touched. Google's spam policies say that sites which violate them may rank lower or not appear in results at all. A suspended Business Profile removes a local business from the place many customers look for it. None of these is predictable from inside the business, and each is easier to survive with a second route already running.
Rising prices. Paid channels are sold by auction. When more competitors bid for the same people, or the platform changes how it sells, the cost per customer rises whether the business changes anything or not.
Ranking, feed and algorithm changes. Organic search rankings, social feeds and marketplace search results are decided by systems the business cannot see. A change can reduce reach overnight, and the business may never learn why.
Marketplace rules. The marketplaces versus own store comparison explains that marketplaces keep the customer relationship, restrict marketing to buyers outside the platform and apply their own rules and penalties. A seller who depends on one marketplace depends on its fees and rules too.
Saturation. Even without outside changes, one channel eventually runs out of new people to reach at an acceptable cost. Each extra unit of spend buys less, the pattern the diminishing returns entry describes, and the guide to plateaued marketing growth shows how a saturated core channel looks in the numbers.
Owned audiences as insurance
The best protection against losing a rented channel is an audience the business owns: people it can reach without asking a platform's permission. The owned, earned and paid media guide explains why a business without owned media has to buy or earn attention first, every time.
- A consented email or SMS list. The most direct form of owned audience. The list-building and consent guide covers sign-up placements, consent records and keeping the list healthy. Track its List Growth RateNet change in an email list over a period as a share of its size at the start. as a measure in its own right, not as an afterthought.
- Customer accounts. Logins, order history and preferences held on the business's own systems are First-Party DataData a business collects directly from its own customers, visitors and users. that no platform can switch off.
- A community. A group of customers who talk to each other and to the business, whether in a forum, an event series or a club, keeps the relationship alive between purchases.
- Brand demand. People who search for the business by name, visit directly or ask a stockist for it by name are not owned in the same sense, but they are far less exposed to any one platform. Growth in Branded SearchSearch queries that contain a company's or product's brand name. is a sign that marketing is building memory rather than renting attention.
Turning paid customers into owned relationships is the single most useful habit for a concentrated business. Every order from the main channel is a chance to ask for permission to stay in touch, and every message to the list is a contact that did not have to be bought again.
Choosing what to diversify into
Diversification is not adding channels at random. The channels overview lists what each channel is for and how it is bought, which is a quick way to see the candidates. Use the method for choosing channels, which locates demand and attention, gives each candidate a job and checks the economics, with one extra question: does this route reach customers by a different mechanism from the main channel?
| Main channel | Diversification that shares its risks | Diversification that reduces them |
|---|---|---|
| Ads on one social platform | The same ads on a second social platform | Search, where people already look for the product; stockists or partners; an owned list |
| Organic search | Paid search on the same search engine | An email list, referrals, a marketplace presence |
| One marketplace | A second marketplace with similar rules | An own store with a consented customer list |
| Referrals from one partner | Another partner of the same kind | Search, content and a direct relationship with end customers |
The second route should do a job the first does not, which is the logic of designing a mix in the channel mix guide. A business living on demand created by social ads, for example, may find that search captures the demand those ads already create, and that stockists reach people who never see the ads at all.
Paying for diversification
A new channel usually costs more per customer while it is being learned. The audience, the creative, the targeting and the measurement all have to be worked out, and the early results are rarely representative. Budget for that learning as a cost, and judge the new channel on the milestones that matter at its stage, not on whether it beats the main channel in its first months.
Two ideas help size the investment.
- The 70-20-10 budget rule sets aside a share for activity that is not yet proven, so that the main channel's visible results do not win every comparison. The framework's own advice applies: work out what each test needs to give a readable answer, and run fewer tests rather than thinner ones.
- Marginal returns in the main channel. If the main channel is near saturation, the last units of spend there may already be buying customers at a poor cost. Moving those units into a second route costs less than the channel's average results suggest.
Time is part of the budget too. Building an owned audience costs hours of writing, design and list care, and those hours need protecting from the main channel's demands.
A contingency plan for the main channel
Diversification takes months. A contingency plan covers the gap, and scenario planning gives it a shape: signposts to watch, triggers at which to act, and responses agreed in advance.
- Name the signposts. Rising cost per customer in the main channel, policy warnings or disapprovals, falling reach for the same spend, a drop in organic visibility, marketplace notices.
- Set the triggers. The point on each signpost at which the business acts, taken from its own figures and tied to the reporting rhythm that already exists.
- Write the first week. If the main channel stopped tomorrow, who would do what? Which customers would be emailed, which second channel would receive budget, which creative is ready, who would deal with the platform, and how much cash covers the gap?
- Protect access and ownership. Make sure the business, not a supplier or a single employee, owns its accounts and holds administrator access, so that a dispute or a departure cannot close a channel. The guide to manager accounts and user access sets out the checks for one platform, and the same principle applies to every account.
- Keep what you are allowed to keep. Customer records, consent records, creative files and reporting history held on the business's own systems survive a lost account.
Review the plan when the main channel changes, and rehearse it once by asking what would happen if it fired today. Dependence on one channel is one entry among several in a marketing risk register, the tutorial that records each risk's warning signs, owner and prepared response and keeps them under a regular review.
A hot-sauce brand loosens its reliance on one platform (illustrative)
Fizzlethorn Hot Sauce builds a second and third route
Fizzlethorn Hot Sauce, an invented small-batch brand selling online, counts where its new customers came from over the last twelve months. Of 4,200 first orders, 3,150 came from ads on one social platform, 420 from organic search, 380 from referrals and 250 from other sources. Its cost per new customer on that platform has risen from £11 to £16 over the year, and it has an email list of only 900 addresses, because checkout never asked.
| Source of first orders | Last twelve months | Share |
|---|---|---|
| Ads on one social platform | 3,150 | 75% |
| Organic search | 420 | 10% |
| Referrals | 380 | 9% |
| Other | 250 | 6% |
The owner asks what would happen if the ad account were restricted for a month: about 260 fewer first orders, with no way to reach past buyers except the small list. The plan has four parts. A consent box at checkout and a monthly recipe email turn buyers into subscribers, with a goal of 4,000 addresses within a year. A search campaign on sauce and chilli terms tests whether people already looking can be captured at under £14 per customer. A stockist programme approaches twelve independent delis, which reach customers who never see the ads. And a one-page first-week plan names who would email the list, which search terms would get the social budget, and the two people with administrator access to every account.
Ten months later, ads on the social platform still bring the most first orders, but at 58% of the total; the list has 3,600 subscribers, recipe emails bring repeat orders every month, search produces new customers at £13 each, and seven delis stock the sauces.
Common mistakes
- Measuring by channel but not by platform. Two channels owned by one company are one dependence.
- Waiting for a shock to start. Diversification takes months; the time to begin is while the main channel still works.
- Diversifying into channels that share the same risks. A second platform used the same way protects less than a route that reaches customers by a different mechanism.
- Judging the new channel against the mature one too early. New channels cost more while they are learned; judge them on their own milestones.
- Taking money from the main channel's efficient spend. Move the least efficient units, not the best ones.
- Treating the email list as an afterthought. Every customer won through a rented channel is a chance to build an owned relationship.
- Letting a supplier or one person own the accounts. Access and ownership are part of the contingency plan.
- Having no written plan for the first week. Decisions made in a crisis are slower and worse than decisions agreed in advance.
