Optimize the business at twenty to forty-nine people

1. The stage

This is level six, and its theme is to optimize.

Item	At this stage
Headcount	20 to 49
Your role	Leader: you were a manager, but you can no longer manage everybody, so you have to lead
Leadership structure	Two full layers of managers: managers who manage managers who manage individual contributors, with you on top. In titles: directors, managers, individual contributors, and you as the one leader above the directors
How many companies get here	Only 2.1 million businesses

It is hard to get here because, in general, you cannot do everything yourself now: the work goes through other people. The early stages' role names are, to a degree, made up, since so small a business has no real titles; for two stages you have been, more or less, above the highest title in the business. Now you lead directors; the next stage will need leaders.

2. Each function at this stage

Function	What breaks	What graduates it
Product	You are improving two products, so product improvements slow down	An incremental product-improvement process for both products
Marketing	Ads convert less well to colder traffic, so the cost to acquire a customer goes up	The ad assembly process, to raise volume, run weekly at a minimum
Sales	Closing efficiency on the core product drops and becomes inconsistent: refunds go down, but so do closes	A sales training system, individual coaching, and a team cadence
Customer service	Churn rises with the volume and the new product	Segment customers by cohort, and create activation points
Information systems (called IT)	New people come in, and old people leave with intellectual property, data and access	Cybersecurity, malware protection and password protection for employees, with onboarding and offboarding procedures
Recruiting	The number and type of roles you need is beyond your current network: everybody good you already knew has a job	Head-hunt for higher-level roles, and add several interviews and interviewers, each filtering for something different
Human resources	You cannot get higher-level talent, because they expect full compensation packages: benefits, retirement and so on	Benefits, retirement, and advanced compensation plans
Finance	You spend money to grow and do not grow; money seems to be missing, and you cannot see the return	More granular financial data

3. Product: one problem at a time, across both products

With two products you are balancing two plates. Put one incremental improvement process over both:

1. List every problem with both products.
2. Estimate each one's gain, cost and risk: what it will make you, what it will cost in time, and how likely it is to happen. You could add how many people it affects, and by how much.
3. Re-rank by highest impact and lowest cost.
4. Fix them one at a time, in that order.

Flag: which problem goes first. Here, problems are re-ranked by highest impact and lowest cost; SOP 164, section 3, goes for the biggest reward, not the easiest problem. Weighing cost clears cheap wins first; chasing reward takes the hardest win first. This page does not settle which reading is right.

This echoes a framework named by the initials of reach, impact, confidence and expense; almost all such frameworks share the same three to four elements. It is a cycle, not a box you tick: resources expand and you re-prioritize. SOP 148 section 3 ranks fixes at an earlier stage.

Fix in sequence. A lot of people want several fixes at once. As a rule of thumb here, you must be able to call something a fire (your conversion rate, sales) and still leave it, because it is not the biggest problem. Put four people on one problem, with the other fires still burning, and they finish it and move together to the next: most businesses hold the magnifying glass too far away and only warm the paper. Schedules clear except for that one problem make the pace addictive, it is said. Sometimes the order is forced: you cannot walk through the building until the door is open. A lot of times it is not, and many problems share one pool of people.

Problems before missed opportunities. A missed opportunity is money you could be making, not money you are losing; a problem exists now. In one practice here, solving problems rather than chasing opportunities made the existing opportunity bigger. Chase the opportunity first and you scale your problems plus the new ones it brings, with fewer resources. Tie things off as you go and you grow by getting better, not by doing more: the theme of this stage.

Let things be imperfect. A second view: wanting ten out of ten in every function at all times creates competing priorities, and a lot of the times such an owner must learn to let things be imperfect. Every solution creates other problems; the only perfect businesses probably exist on paper, go under, or make no profit. Perfect everything at once and everything becomes mediocre.

Choose features two ways. Features are what you bundle into what you sell.

• A community vote. The view offered dislikes it and calls it middling, though not terrible: customers judge from what they know, so where you could do things they do not know you could do, you miss the big jumps. The view here is that it works better where technology is minimal; the most innovative understand their business more deeply than their customers do.
• Prune. Liked a lot. List every feature and ask: if all went but one, which would you keep? Then: if all stayed but one, which would make no difference to your life?

Overwhelm is said to be the top reason people cancel memberships and subscriptions. Better, not more, is better, and you often get there by pruning, as a gardener prunes a tree.

Delete quietly. With the top and bottom two or three known, delete the bottom one and say nothing. Asking first upsets everyone, because you are taking something away; yet if all ranked it lowest, nobody is using it, or if they are, maybe they get no value from it. Then wait. If nobody says anything, you make more for doing less, with a better product. If people do respond, before adding it back ask whether they are the customers, and the part of the base, you want to grow. Sometimes the cut is strategic: you accept losing 5 percent of customers, in the example given, for a healthier whole. SOP 131 asks the same pair of questions twice a year.

Flag: when people miss a cut feature. Here you weigh whether they are the customers to grow before adding it back; SOP 168, section 3, has the few who insist pay more for it. Weighing first can cost some customers for a healthier whole; charging the few keeps their money and a simple core, with an add-on to run. This page does not settle which reading is right.

4. Marketing: assemble ads at volume

You now advertise on more platforms and spend more a day, and ads fatigue faster: you need more ads and more variety. Earlier, lifetime value was too low to convert; now acquisition cost keeps rising. It feels similar, but the fixes differ.

The assembly is SOP 119 — Assemble hook, meat and call to action: 50 hooks, three to five meats and one to three calls to action, 150 to 750 variations a week, 80 percent of your time on hooks. The reason given: an old advertising saying that once the headline is written, 80 cents of your dollar is spent, read as only 20 percent of people getting past the hook. You could run one meat and recycle hooks to reach new people; fresh hooks prevent banner blindness.

Hooks, in order of risk: your past ad winners, your content winners, other people's ad winners (a favorite is unrelated industries: a hotel chain's hook could work with your meat), other people's content winners, and last the public ad libraries most platforms keep, where an ad running for a month or two is probably a winner. With no winners yet, make up 50. Joining the first part of one hook to the second part of another is called the magic, but habitually the other places come first: the chance of a past ad winner working again is called super high, of a content winner on the same audience pretty high.

Flag: which place to find hooks is likeliest to work. SOP 119 (section 3) gives the same places in the same order, but says hooks from your free content have the highest likelihood of converting, and that it is hard to tell from an ad library which ads perform. Here, past ad winners are likeliest to work again, content winners second, and an ad running a month or two is probably a winner. This page does not settle which reading is right.

Meats. A meat is, more or less, another way of stating the offer. The habit described shows each of the four value drivers (SOP 3) from both sides: more good stuff, less bad stuff.

Driver	The good side	The mirror image
Dream outcome	The good stuff	The nightmare, like a black-and-white ad about what goes wrong
Risk, or perceived likelihood	The low risk of taking your option	The risk of not doing this
Speed	Your way is faster	Other ways are slow and delayed
Effort and sacrifice	How easy and sacrifice-free yours is	The effort and sacrifice of other ways, or of not using your solution

The five themes named for the meat (the product or service; the people who have done it, as testimonials and sometimes employees; education; narrative and emotion; non-human, such as cartoons and memes) run close to SOP 120. When showing the product or service, show the result people want: an agency could show leads, but its clients want a business full of customers. People buy the drill to get the hole.

Calls to action say what to do and how, when, what they get, and what happens next, with a demonstration. The habit given is to show the funnel's next page at the end of almost every ad, so it is congruent; that can lift all ads significantly (SOP 119 section 5). What is probably limiting your advertising is not your market but ads that are not good enough; the coldest traffic needs much better creative. Dividing the hook work itself is on SOP 122.

5. Sales: train, coach, keep a cadence

You probably have a team now. The frame preferred is that no one is perfect: tell the team you are no better than they are and will role-play and mess up with them. Feel you must look perfect and you will not role-play enough, for fear of looking stupid.

1. A sales training system. Recordings of really good sales and of the perfect script, and a tone guide on the script: italics, underlining, bold and capitals mark where to slow down, speed up, ask, or ask with a question frame ("You like dogs?" is not "You like dogs"), and where the voice lifts or drops.
2. Individual coaching. Role-play each new rep's specific weakness. The practice described typically does this twice a day in week one, once a day in week two, then weekly once the rep is productive. A rep not productive by the end of week two is strongly considered for cutting.
3. Team training. Keep the team-wide game tape and add game tape just for sales. The rep first says what they could have done better; then the team says what they did right, for public praise. The reverse, the team listing every mistake, was found the hard way to be horrible. What comes out shapes the manager's training with that rep: missing rapport means training rapport until it is right.
4. The cadence. A daily huddle and a weekly one-on-one, once the team is at maintenance: in the beginning you onboard this way, and over time you settle into maintenance.

A weekly cross-team call review is SOP 135.

6. Customer service: activation

Customer service is how you get customers to refer, retain, review and resell: the four Rs. Segmenting customers by cohort is on SOP 158; activation is here.

Do not make canceling hard. Hiding the cancel button does not solve churn, and you are promised bad reviews: people hate not knowing whom to email or whether anyone will answer, or having to phone and leave a voicemail. Prevent churn before it happens instead.

Activation is getting someone to actually use what you sell; it is called the biggest correlator to reducing churn. Five steps, each named with the same letter:

1. Map the journey. Outline the key milestones from onboarding to ongoing delivery: what the customer does and what your team does for the customer to realize value. You might call after the sale, onboard on a call, send emails, give a free guide; the customer might start measuring their numbers, running ads and learning to record them. Often, in one practice here, the two journeys are color-coded: it might be the customer's in blue and yours in yellow.
2. Measure patterns. Find which behaviors of your best customers correlate with the highest retention and engagement. An easy way: walk them through the journey in an interview and ask what they do, or skip, that is not on it. A lot of times they do some things differently; in one business, a focus group of best customers showed they often did a few things differently from the rest, and those could then go into the journey.
3. Make a hypothesis. Pick the actions or milestones that could be the activation point. If eight out of 10 of your best customers do two things the others do not, those two might be it. In one business, customers who brought the operator of their facility onto the sales and onboarding calls were almost twice as likely to stay, and to stay longer. It could be that simple.
4. Monitor results. Test whether the milestones drive engagement, conversion and retention: require the two things (everyone brings their operator onto three calls, say) and watch churn for the next 90 days; maybe, if people churn at week six, allow three or four months. Change more than one or two things (more than one, in the view here) and you will not know what worked. Test one thing at a time: not knowing what creates success is called worse than not knowing how you fail.
5. Modify the onboarding and the offer. Rework the journey to hit the real activation point, and build it into your offer and guarantee, if that's possible.

A financial software company gave two months free for watching five live webinar videos five weeks in a row and filling in a survey; one customer stayed about four years, having learned the software there. An offer can tie a rebate to such actions and advertise the net cost after it, as long as you disclaim that it is after rebate. That advertising rule is carried as stated; the law is not covered on this page.

Another route to the activation point, starting from the customers who left, is SOP 129.

7. The functions carried elsewhere

• Information systems: employee security, onboarding and offboarding are on SOP 155.
• Recruiting: head-hunting and the four-interviewer process are on SOP 163.
• Human resources: choosing benefits is SOP 159; pay plans for higher-level people are SOP 160.
• Finance: SOP 154.

8. The bottom line

The constraint is time and resources wasted on a growing workload: more of everything, more leads, more products. To graduate, grow by doing everything better rather than doing more: streamline, focus, and put resources where they return most. The next stage is 50 to 99 people, theme categorize.

The stages are rules of thumb. You might be at level nine for advertising and level three for human resources; what lags will probably become a larger risk as the rest gets ahead. Look back and tie it off, then look ahead and prepare, so you do not backtrack.

9. What this page does not decide for you

• Whether "what happens next" is optional in a call to action. Not established on this page.
• Whether the whole sales team also meets weekly. Not established on this page.
• Two changes at once, or one. The example requires two things; the rule is one at a time. This page does not settle which reading is right.

10. What this page does not cover

Job descriptions and résumé review are SOP 148 section 8.