Section 1
Start with what the buyer actually sees
An office manager at a 60-person dental group has been told to replace the IT provider. She has three tabs open. Here is roughly what the first one says above the fold: "Enterprise-Grade Managed IT Services. Proactive 24/7 NOC monitoring, next-generation EDR, and a fully managed RMM stack backed by certified engineers. Cloud, security and compliance solutions tailored to your business." She does not know what a NOC is. She does not know what EDR is. She knows the last provider also said proactive. Nothing in that sentence can be checked, argued with, or repeated to the practice owner who will actually sign. Here is the same firm, rewritten: "IT support for dental and medical practices with 20 to 150 staff. When something breaks you call one number and reach a named engineer, not a queue. We answer within 30 minutes during business hours. If we miss that, we take 10 percent off that month's invoice, and you do not have to ask. Our first 30 days with a new practice are published here, day by day, including what we will need from your current provider." Notice what changed. The second version is not simpler because simple is nicer. It is simpler because every sentence now contains something a bad provider would find expensive to say. A weak firm can type "proactive monitoring" for free. A weak firm cannot comfortably promise 30 minutes with a refund attached, publish an onboarding sequence that a client can hold it to, or name the human who picks up. That is the whole mechanism of this article. Not tone. Not readability scores. Cost. The question to ask of every line on your site is whether a competitor who is worse than you could write it just as easily. If the answer is yes, that line is decoration.
Section 2
Jargon is a cheap signal, so it separates nobody
There is a name for the situation your buyer is in. Information asymmetry: one side of a trade knows the quality of the goods, the other does not, and the side that does not still has to choose. Your prospect cannot inspect your patching discipline, your alert triage, or whether your 2am escalation is a person or a voicemail box. They will not find out for a year, and possibly not until something fails. When buyers cannot verify quality directly, they read signals. A signal only carries information if it is cheaper for a good provider to produce than for a bad one. That cost gap is the entire mechanism. If a weak competitor can produce the same thing at the same effort, the signal tells the buyer nothing, no matter how impressive it looks to you. Technical vocabulary is the cheapest signal in this industry. It costs nothing to write. It requires no operational capability behind it. A two-person shop that outsources its night coverage to a partner it has never audited can put the identical paragraph on its homepage in an afternoon. Certification badges sit close behind: the exam is buyable, the badge is a footer image, and nothing about it reports how the firm behaves on a bad day. The result is what economists call pooling. Everyone produces the same signal, so the signal stops separating anyone, and buyers fall back on the one dimension they can compare without expertise: price. If your sales conversations keep collapsing into a per-seat comparison against a provider you know is worse, your copy is a contributing cause. It told the market you were the same category of thing, then invited a price test to break the tie. Costly signals are the way out of that. Not louder claims. Claims that would hurt to make if they were not true.
Section 3
Language a buyer cannot parse increases risk, not authority
There is a second effect stacked on top of the signalling one, and it works at the level of the sentence. People experience some text as easy to process and some as effortful. That feeling of ease, sometimes called fluency, does not stay attached to the sentence. It gets attributed to the thing the sentence is about. When a non-technical reader hits a paragraph they cannot follow, they rarely conclude "I lack the background for this." They conclude something vaguer and more damaging: this feels complicated, complicated things go wrong, and I will not be able to tell when it is going wrong. For a buyer who has been burned before, that discomfort has a ready explanation waiting for it. The last provider used language she could not check, used it again in every ticket update she could not evaluate, and then sent a renewal with a number she could not challenge. Your unparseable homepage is not neutral to her. It is familiar. There is also a mechanical problem. The person reading your site is often not the person who signs. The office manager builds a shortlist, then explains it to the owner in a five-minute conversation you are not in. Whatever she cannot restate in her own words does not survive that conversation. "They do next-gen endpoint detection with a managed SOC" will not be repeated. "If they do not answer in 30 minutes we get money back, and they only work with practices our size" will be repeated almost word for word. So the practical standard for MSP website copy is not reading level. It is transmissibility. Can a non-technical reader carry your differentiator across a room without your website open? If not, you are absent from the shortlist conversation, whatever your engineering is worth.
Section 4
The rewrite method: situation first, stack second
Three rules, applied in order. They are unglamorous and they work on any page. Rule one: lead with the situation, not the stack. Open every important block with a moment the buyer recognises from their own week, then attach your capability to it. Not "multi-layered endpoint security" but "a staff member clicks something they should not have, on a Friday afternoon." The buyer cannot judge your security architecture. They can absolutely judge whether you have understood their Friday. Rule two: plain English first, the acronym in brackets after. Never the reverse. "We keep a second copy of your data off-site and test that it restores every month (providers call this BDR)." The bracket does two jobs at once. The non-technical reader gets the meaning without stopping, and the technical reader, or the search engine, or the model summarising your page for someone, still sees the term. You lose nothing by defining yourself. You lose readers by assuming. Rule three: one claim per sentence, and every claim points at something checkable. A number, a named role, a document the buyer can open, or a consequence you accept. "Responsive support" fails. "Answered within 30 minutes, or 10 percent off the month" passes. "Smooth onboarding" fails. "A 30-day onboarding plan you can read before you sign" passes. Then test it the only way that counts. Read the page aloud to someone in your target industry who has no IT background, and ask them to tell you what you do and why you are different. If they hesitate, the copy is not finished. If they hand it back to you in technical words, it is not your argument they are repeating, it is your vocabulary, and vocabulary does not close deals. For how this interacts with page structure and layout rather than sentences, our notes on web design for IT services and MSPs sit at https://bizgrowthaxel.com/blog/web-design-for-it-services-and-msps/.
Section 5
What to cut from a typical MSP homepage, and what replaces it
Work through your own site with two lists. Cut, or demote well below the fold: the certification wall in the hero. The vendor logo bar used as primary proof. "Proactive, not reactive." "We become your IT department." "Tailored solutions for your unique business." "Enterprise-grade." Any capability list longer than five items. Any sentence whose subject is your technology rather than the client's situation. None of these are lies. They are simply things every competitor can produce at zero cost, which means they do no work on the page. Put in their place, roughly in this order. Who you serve, stated narrowly enough to exclude people. Sector, staff count, region. Exclusion is itself a costly signal, because a firm chasing every deal cannot afford to publish limits. A response commitment with a consequence attached. The consequence is what makes it a signal. A response time with no teeth is a wish. A named escalation path. Who picks up, who picks up if that person does not, and what happens at 2am on a Sunday. Roles and first names beat a generic support address. A published onboarding sequence with days on it. What happens in week one, what you need from the outgoing provider, what breaks and when. This is the most under-used asset on MSP sites, and it speaks to the fear that stalls most deals, which is not price but transition. One real engagement told with dates and numbers, including something that went wrong and how it was handled. A case study with no friction in it reads as marketing. A case study with a bad Tuesday in it reads as a record. An engagement floor. The smallest client you take, in seats or monthly value. It disqualifies the wrong buyer before you have spent six hours scoping them. Exit terms. What happens to their data and documentation if they leave you. Almost nobody publishes this, which is exactly why it separates.
Section 6
The honest cost of costly signals
The name of the mechanism should warn you. These signals work because they cost something, which means you will pay. A response commitment with a credit attached will be missed sometimes, and you will pay out. That is not a failure of the design, it is the design. If the payout would be ruinous rather than annoying, your number is wrong, not the approach. Set the window at something your dispatch already achieves on a bad week, not on an average one. A published promise you break in month two is worse than the jargon it replaced, because now the buyer has evidence rather than doubt. Naming humans creates a maintenance problem. People leave, and a site listing an engineer who resigned in March reads as neglect. Budget a quarterly pass over those pages, or name roles and shift structures instead of individuals if your turnover is high. Publishing your onboarding sequence hands competitors a document to copy. Accept it. They can copy the page in an hour and cannot copy running it, and if copying it forces them to actually deliver a structured onboarding, the buyer is better off and you are still ahead on execution. The advantage you are protecting is operational, not editorial. Narrowing who you serve will cost you inbound volume, immediately and visibly, before it pays back in close rate. If your pipeline is thin right now, that lag is a real risk, and the right move is to shrink the scope of the change rather than pretend the cost is not there. Start with one page and one segment. The largest cost of all: none of this is copywriting. Every item on the replacement list is an operational commitment written down. If the operations are not there, do not write the sentence. Build the escalation path first, then publish it. Our broader working notes on MSP acquisition sit at https://bizgrowthaxel.com/msp/, and the sequencing point holds across all of it. Publish nothing you cannot survive being held to.
Section 7
Where this inverts, and what the model cannot see
Segment before you rewrite, because there is a buyer for whom every argument above runs backwards. If the prospect has in-house technical staff, an IT manager, a sysadmin, a developer who inherited the servers, the evaluator is no longer non-technical and the asymmetry narrows. That person can check your claims. They will ask about patch rings, conditional access policy, how you handle a failed restore test. Against that reader, technical depth becomes the costly signal, because a weak provider cannot sustain a coherent technical answer under follow-up questions, while anyone at all can say "we answer in 30 minutes." Vagueness now reads as evasion. So the segmentation question is not company size, it is whether internal technical capability exists. As a rough guide, firms under about 50 seats rarely have it, firms over 150 usually do, and the middle is mixed. The practical answer is two paths, not one blended page. A business-owner path in plain English, and a co-managed path with real technical specificity for the internal IT lead, each linked clearly enough that readers sort themselves. What you must not do is average the two, which produces copy that is too technical for the owner and too shallow for the engineer, and signals nothing to either. Now the blind spot, and it is a large one. Both lenses used here, signalling and processing ease, assume a buyer who arrives with no prior information about you. In a local market that assumption often fails. An MSP that has been in one mid-sized town for 15 years, whose owner coaches the league and whose clients all know each other, is being chosen on accumulated reputation that no homepage can move. Copy is downstream of memory there. This cuts both ways. An incumbent may rewrite the whole site and see nothing change, because the pipeline was never running through the site, while a firm entering that market gets far more from identical work. Before you invest, ask your last 20 leads how they found you. If the answer is overwhelmingly "someone told me about you," fix the referral engine first and treat the rewrite as maintenance.
Section 8
The fitness test
You are ready for this rewrite if the following are true. You can name the buyer segment you want and the one you keep taking by accident, and you are willing to publish limits that exclude the second group. You have an escalation path that already exists in practice, with a person at the end of it, not an aspiration in a proposal template. You know your real response times on a bad week, not the number in your service agreement. You have run enough onboardings that you would be comfortable describing one publicly, including where it got rough. You have at least one client who will let you publish dates and numbers. And you have the tolerance to watch inbound volume dip for a quarter while close rate is the thing that moves. You are not ready if any of these are true. Your operations cannot currently keep the commitments you would be publishing, in which case the rewrite converts a vague problem into a documented one. Your pipeline is majority referral and word of mouth in a market where you are already known, in which case this is not your binding constraint. Your buyers mostly have in-house IT, in which case read this article inverted and go deeper technically rather than plainer. You need leads this month, because signalling changes work on the deals that arrive after the change, not on the ones already in motion. Or nobody in the business has the authority to promise a consequence, in which case you will produce another page of adjectives with a slightly better rhythm. The test in one line: if a worse competitor could copy your homepage tonight and lose nothing by it, you have not written copy. You have written the category.