Audit delivery by who values what, then cut its cost

1. Where this move sits

Lowering the cost of goods sold, what it costs you to deliver, is the next of the eight ways to make a customer worth more, in the order used here. The eight as an exercise are on SOP 232 — Drill the eight ways to raise what customers are worth; the list itself opens SOP 72 — Generate tiers and downsells from the quality vectors. This move opens with a survey, called a really good tactic, and follows it with a run of cuts.

2. The survey: three questions

Consider sending this survey to your customers. It has three parts, in this order.

Which customer this is. Ask for the customer's name and, if you sell business to business, their average revenue. Otherwise it could be their address, or whatever else tells you who they are. You want to know which customer gave each answer; section 3 is the reason.

Last one standing. Here is the core question. Ask them to picture you being forced to take away every component of what they bought (the plan, the service, the offer, or whatever it is) except one. Which one would they battle hardest to have you hold on to? That is the very last thing they would let you cut.

First to go. Then the flip side. Ask them to picture you cutting costs so that you could take away just one thing from everything they get now. Which one would they be gladdest to see disappear the next day, the thing they would not miss at all?

The two component questions are called very valuable ones that software-as-a-service companies ask, and are said to work 100 percent for a service business too. The reason given: software as a service is a service, so what works there works in service.

In one demonstration, the components of one offer were listed and people answered by letter, so they could see how it works; a more formal version was probably to follow.

The same pair of component questions, in the same order, is section 2 of SOP 131 — Run the keep-one and remove-one survey, which also sets how often to send it. What this page adds is the question about who is answering, and what to do with it.

3. Read the answers by who gave them

Do not weigh every customer the same. That is why the survey opens by asking who is answering. Suppose your poorest customers, the ones who give you the most trouble, all want thing one, while your rich customers all want thing two. Then the view here is to want more of thing two and to get rid of thing one. It is not purely a matter of doing whatever everyone says.

An even spread is good news. If the votes spread evenly across your components, you made a great offer: each component you added won over the particular people it was there for. This ties to offer components, which are ground for SOP 230 — Shape the front-end offer for lifetime value. With a perfect split, one sixth to each component in the example, the view here is to leave the offer exactly as it is.

A cluster tells you where to spend and what to cut. If you are trying to bring costs down, find what people value most and give it more of your time. Then look to cut, if you see the answers pile up on one component, sometimes two.

Check who is voting to cut. Before you drop something people say is not valuable, ask whether those people are your good customers or your bad ones. This is a little bit of nuance: the survey is not simply a voting machine, and you check which people are giving which answer. Beyond that, that is the game.

4. The department nobody missed

One company asked the first-to-go question. Here is what happened.

Too much on the list. Written out, everything that business delivered made a list far longer, it is thought, than the offer in the demonstration above. The owner was earlier in his career and did not yet see that the aim was fewer, better things; he kept over-delivering and was wearing himself out trying to keep everyone happy.

What customers valued. The answers did cluster; this applies to that company. Lead generation and sales were the top two, and that is what those owners wanted. In this experience, the smaller the business owner, the more common that is.

What they would drop. On the first-to-go question there was a massive cluster around one department the owner had built to help customers, separate from customer support. A lot of effort went into it.

The result. The survey named that department first to go, overwhelmingly. The owner listened to the customers and cut it. Churn did not change at all.

The cost no longer carried. The staff, the managers over them, the headaches, the HR work, and training that was a pain to run. None of it had mattered.

The lesson. Your business has components right now that take a lot of effort and make little difference. That is what makes the survey valuable.

5. Other cuts to make now

Other ways to lower what your services cost right away follow. They are just some tactics, in the hope that one of them will apply to you.

• Serve more customers per person. You can push up the ratio of employees to customers, as it is worded. If you work one-on-one, think about semi-private sessions; move from one-on-one toward one-to-many.
• Offshore talent. Anything that need not be done in person can be done offshore. It is claimed that you can hire people with doctorates for 20 dollars an hour. One owner's own picture of offshore talent was, that owner thinks, more skewed many years ago, and some exceptional people there, it is said, work extremely hard. The view here is that there is a lot of value in it if the work does not have to be in person.
• Tier the delivery. You can also tier who does what. Say an expensive person now does four things. Maybe two of them could go to someone who is not expensive, so that you can get twice the leverage from the expensive person.
• Sell to more similar customers, so you can productize delivery. You can sell to more customers who resemble each other. It is seen as a major lever for many owners, in particular those with revenue between "$1 and $3 million", a band then widened: "Call it $750 and $3 million". At that one-to-three-ish size, it is typically true that your customer avatar has to get tighter and more disciplined. The reason, when the aim is lower cost and a higher gross margin for the business or the offer: similar customers have similar problems, so your solutions can become products. It is still a service, but you can templatize a large share of it, maybe 80 percent repeated with a very small part custom, rather than selling to everyone. It is counted among the biggest places to improve for most businesses of that size, and said to be usually why they get stuck in operations. The way out is to make it easier on yourself: turn down the people who will not pay you as much, take on the ones who will, and have them be alike. Whatever label you like will do: narrowing down, choosing an avatar, niching down. Done this way, you can really bring delivery cost down: the customers spend more and cost less to serve. Choosing a richer customer is the ground of SOP 229 — Sell to richer customers and rewrite what you measure.
• Move to done with you, from done for you. You can also make this switch. A simpler one, and one many people are totally up for.
• Guardrails on project work. If you deliver done-for-you work, especially anything construction related or where you build things, set limits: the job cannot exceed these units, this time, "this whatever". That helps bring cost down. One really elegant version, especially in home services, though in this view it works for anything: you can guarantee the timeline or the cost (or whatever you promise), a promise that holds only if they keep within the guardrails. A customer may choose to go outside them; that is fine, but the guarantee is then void, and so is the budget you gave them. The reason given: the front end can still be sold hard, and once customers change their minds, "which they all do," the back end can still take as much selling as you like.
• Automate portions, not the job. This should be obvious. People get lost trying to automate a whole job. The encouragement here is to break it into chunks instead: if you split a job into 100 steps, each step is far easier to automate than the full job. It need not be your own job; you can break your employees' work into chunks the same way. In this view, it is to be counted as the high-leverage work you do one time that makes the company and its products more valuable for good.
• Batch the work. Service businesses do this a lot, if you have the opportunity: everyone does one type of work in a batch on Fridays, say, or one department takes one kind of work on a set day. On its day, that team works more efficiently, which lowers costs.
• Synchronous to asynchronous. This one is probably more for consulting or business-to-business work. If you have a touch point of an hour a week, or whatever it is, you can replace it with an eight-minute recorded video message or a Slack update, or whatever suits. Typically, it is said, the change saves your team 60 to 80 percent of its time, and customers like it.
• Audit delivery. The survey above, as the last of the run: take out what people do not care about.

6. Beside SOP 127

SOP 127, the page on cutting delivery cost, carries its own cost levers. Where a lever here is the same, that page's version is not repeated; the right-hand column is what this page adds.

Here	On SOP 127	What this page adds
More customers per person	Lever i, where an account rep is given more clients	Semi-private sessions, and one-to-one toward one-to-many
Offshore talent	Lever ii, with a savings figure of its own	The in-person test, and the claimed 20 dollars an hour
Tier the delivery	Not on that page	The whole lever
Similar customers, productized	Lever iii, with templates and automation	The revenue band, maybe 80 percent templatized, and turning some buyers down
Done with you in place of done for you	Lever iv	Nothing further
Guardrails, and a guarantee that holds only if they keep within them	Lever v caps use and charges past the cap	The guarantee and the budget, void outside the limits
Automate portions	Named under lever iii	Breaking the job into steps first
Batch the work	Not on that page	The whole lever
Synchronous to asynchronous	Lever viii clears out recurring meetings on a schedule	Recorded updates in place of live touch points, typically saving 60 to 80 percent of team time
Audit delivery	Not on that page	The survey, read by who answered

Lifetime to annual access, remote delivery and bulk buying are on SOP 127 and not on this page.

7. What this page does not decide for you

• The unit of the $750. The first band runs from "$1 and $3 million"; it is then widened to "$750 and $3 million". The 750 is given with no unit. Not established on this page.
• How much weight each group gets. Where the two groups differ, the example builds more of what the rich customers want and gets rid of what the poorest, hardest customers want. This page gives no figure for how far one group's answers should outweigh another's.
• How often to run the survey. This page gives no figure for how often to run the survey. SOP 131 gives a frequency.
• Where to set the guardrails. Units, time, "this whatever": which limits suit which job is not established on this page.
• The void guarantee. Voiding the guarantee and the budget outside the guardrails is one owner's tactic, carried as that. Guarantee law and contract law are not covered on this page.
• Hiring offshore. The pay figure and the view of offshore workers are claimed, not shown. Employment law and cross-border hiring rules are not covered on this page.

8. The checklist

Step	What to do
Ask who	The customer's name, and average revenue if business to business; otherwise it could be their address
Ask what they keep	If every component went but one, which one they would battle hardest to have you hold on to
Ask what they drop	If you could take away just one thing, which one they would most gladly lose
Read by customer	Weigh answers by who gave them; in the example, more of what the rich customers want
Read the spread	An even split means a well-made offer; look to cut where answers pile up on one component, sometimes two
Check the voters	Ask whether the people saying something is not valuable are good customers or bad ones
Cut	Take out what people do not care about
Other cuts	Any that apply, as tactics: serve more per person; offshore what need not be in person; tier the delivery; sell to similar customers; move to done with you; set guardrails; automate portions; batch the work; go asynchronous

9. What this page does not cover

The offer's components as something to sell are on SOP 230 — Shape the front-end offer for lifetime value, and getting customers to buy again is SOP 235 — Get customers to buy again: make it good, onboard, remind. Trimming an offer down to what is high value when you first build it is section 3 of SOP 5 — Cut costly pieces from an offer and bundle the rest. For a gross margin goal and three ways to close the gap to it, see SOP 226 — Set a gross margin goal and close the gap three ways.