About half of your 28.6% admin tax is just tool-hopping. That is the fragmentation tax, and it is separable from the work itself.

Fragmentation is the share of your 28.6% admin tax that exists not because the work itself is large, but because the same client context (a transcript, a deal stage, an hour, a draft) has to be hand-carried across 6 to 9 disconnected tools. The work itself is small. The cost is the carrying. On a typical solo practice running CRM plus invoicing plus PM plus notes plus time tracker plus email plus scheduling plus docs plus knowledge base, fragmentation accounts for roughly half of the admin tax, or 4 to 5 hours a week. Concretely: a 45-minute Zoom call generates around 6 minutes of write-content (the action items, the recap, the time entry, the next task) and 25 to 30 minutes of move-content-between-tools. The fragmentation tax is the second number.

Tool sprawl is the count: how many subscriptions you pay for. Fragmentation is the cost: how often the same fact has to be re-typed into a different system because the systems do not share it. You can have nine subscriptions and low fragmentation if those nine systems share state through one execution layer. You can have three subscriptions and high fragmentation if those three are walled off from each other and you are the integration. Counting subscriptions misses the actual problem; the right metric is writes per client interaction.

They solve some of it. The well-trodden integrations cover roughly the top 30 percent of writes between the most popular pairs (Fireflies to HubSpot, Calendly to most CRMs, QuickBooks to most banks). The remaining 70 percent is either pair-specific (your CRM is Folk and your invoicing is FreshBooks; the integration covers contact sync and not invoice line items), shape-specific (the integration moves a transcript object but not the action items you actually want as CRM tasks), or content-specific (no integration writes a tone-correct follow-up email body, only an empty draft). Native integrations cap out at the field-mapping layer, and the fragmentation tax lives above it.

Two columns in a text file for one week. Column A: every block of admin time over 5 minutes, what tool you were in, and what client interaction triggered it. Column B: out of that block, how much was reading or writing the actual content (the action items, the email body, the line item) and how much was navigating, switching, copying, or re-entering. At the end of the week, the column-B sum divided by the column-A sum is your fragmentation rate. Most solo consultants come in between 45 and 60 percent. The reason it surprises is not the total, it is the per-Zoom-call detonation: a single 45-minute call almost always shows up as 25-plus minutes of column-B time before any thinking happens.

The CRM, by a wide margin, because three other tools (notes, PM, email) have to write into it after every interaction. The CRM is also where consultants are most reluctant to switch, because deal history and contact log compound. The second largest is invoicing, because the time tracker and the PM tool both need to feed it on a milestone or month-end cadence. The third is the inbox, because every other tool either drafts into it (notes app drafting a follow-up) or reads from it (CRM detecting a reply). The 9-category map below names all nine, but in dollars the top three categories produce around 70 percent of the fragmentation tax.

It eliminates the fragmentation between the line items that suite owns end to end (typically proposals to contracts to invoicing on one cadence). It does not eliminate the fragmentation between the suite and the things that stay outside the suite (your meeting transcripts in tl;dv, your inbox in Gmail, your CRM if you do not migrate it, your PM tool if your team uses Notion). For a one-shape practice (wedding photographers, fixed-scope brand work) the suite cuts the tax cleanly. For a multi-shape practice (advisory plus implementation plus retainer, on different cadences with different stakeholders) the suite covers two of nine categories, and the other seven still produce most of the original fragmentation. The honest test is whether the suite owns end to end at least four of your nine categories; below that, you have moved the tax, not removed it.

Clone takes plain English instructions and drives the apps already on your screen. The CRM stays in HubSpot or Pipedrive or Folk. The transcript stays in tl;dv or Fireflies or Granola. The invoice stays in QuickBooks or FreshBooks. The follow-up draft lands in Gmail. The instruction layer is where the seven writes after a Zoom call collapse into one execution, instead of seven manual switches across seven UIs. It does the moving-content-between-tools part of the fragmentation tax. It does not touch the human-judgment part (rate review, contract negotiation, sensitive followup). $49 per month solo, $129 per seat boutique, free 14-day trial, runs locally on your machine so client data does not leave your computer.

Three structural reasons. First, the in-house team has someone whose job is the moving-content-between-tools work (an ops coordinator or sales-ops analyst); the solo consultant is that person and the practitioner. Second, the in-house team usually runs one CRM and one invoicing tool by company policy, so the integration paths are well-trodden; the solo consultant runs whatever each client expects, which fans out the pairs. Third, the in-house team has someone on call when an integration breaks; the solo consultant either fixes it themselves (more fragmentation, just admin instead of execution) or lives with the broken state (which produces stale CRM data and late invoices). The 28.6% admin tax is solo-specific because the carrying cost has nowhere to go.

The 28.6% admin tax math, the line-item bill, and the reducible-versus-irreducible split are at /t/billable-hours-admin-tax. The four mechanical shapes of automation (Replace, Glue, Delegate, Drive) and which one closes which loop are at /t/automate-consulting-back-office. The two-clock model (drafting decays four times per day, approving decays zero) that lets the reducible loops come off your Friday block is at /t/batch-consulting-admin-work. The post-call admin pattern (the single largest fragmentation source) is at /t/post-call-admin-automation.

It is a real category once you separate write-content from move-content. Time-and-motion studies of professional services workflows have measured both numbers since the late 1990s; the move-content number is consistently 40 to 60 percent of total back office time across consulting, accounting, and law practices. The framing as a tax is a marketing choice (the same way "opportunity cost" is a framing). The underlying number is the same number a CFO would call "non-value-added time" in a Lean review: time you paid for that did not produce a deliverable, did not move a deal, did not bill a client. Fragmentation is the dominant source of that number for solo practices.

Pick the single tool category that produces the most writes after every client interaction (almost always the CRM). For two weeks, log every CRM write that came from another tool (a transcript, an email reply, a calendar event, an invoice paid). At the end of two weeks, you have the actual fragmentation between your CRM and the rest of the stack, in writes per week. Then ask: which of those writes could a plain English instruction (from the transcript content, from the email content, from the calendar content) have generated as a draft, leaving you to approve? The answer is usually 60 to 80 percent of them. That is the cut you are sizing.