DEV Community: Natraj L

The Freelancer's Year-End Close-Out: Six Steps to Do Before January

Natraj L — Sun, 18 Oct 2026 14:11:56 +0000

Every December, most freelancers do one of two things: panic about taxes or coast until January. Both waste the most useful data you own: twelve months of real numbers about who paid you, how fast, and how much of your life each client consumed.

1. Close the books

Pull eight numbers: invoiced, received, outstanding receivables, expenses, billable hours, non-billable hours, revenue per client, and your best and worst months. Then compute three things:

Net margin = (cash received - expenses) / cash received
Concentration risk = largest client revenue / total revenue. Above 40% means one email can cut your income nearly in half.
Collection gap = invoiced - received

2. Audit your clients

Score each client 1 to 5 on pay, promptness, scope, respect, growth and fit. Sort them into keep, fix or fire. Use your effective hourly rate for the pay score: invoice total divided by every hour the client consumed, including calls and unpaid revisions. A client who looks like $60 an hour on paper can be $42 once you count the extras.

3. Fix your pricing

Work backwards: income you want + business expenses + tax and retirement set-aside, divided by the billable hours you can realistically sell (most freelancers sell 900 to 1,200 a year, not 2,000). That is your minimum hourly rate.

4. Raise rates

Give 30 to 60 days notice, state the new number plainly, name the effective date, and offer one option such as a retainer. Do not justify the raise with your costs. Expect to lose about one client in five, and make sure it is one from your fix or fire list.

5. Rebuild the pipeline

Use 10-5-2: write down 10 names, send 5 personalized messages this week, aim for 2 real conversations. Past clients are the cheapest source of work.

6. Plan 2027 by quarter

Set one income target, then give each quarter one theme: reset prices and clients, fill the pipeline, deepen the best work, protect margin and rest. Add a weekly cap on billable hours before you need one.

Quick win: filter your invoices for anything 30 or more days overdue and send a two-line nudge today.

I turned this process into a 13-page working guide with the scorecard, pricing worksheet, three paste-ready rate-raise emails and quarterly planning pages: The Freelancer's Year-End Close-Out & 2027 Plan ($12 on Gumroad, $9 on Etsy). I also send one practical template or script a week at AI Founder Weekly.

Originally published on Medium.

How to Write a 30-60-90 Day Plan Your Manager Will Actually Trust

Natraj L — Sun, 11 Oct 2026 13:48:55 +0000

Most 30-60-90 day plans fail for one reason: they list activities instead of outcomes. "Meet the team" is not a plan. "Know who owns each decision on my project by day 30" is.

Days 1 to 30: Learn

Name what you will understand, not just who you will meet: the product, the customers, how decisions get made, and the one metric your team is judged on. End the phase with a short written summary of what you learned and what surprised you.

Days 31 to 60: Contribute

Pick one small, visible win that fits what you learned and ship it. A first fix, a clearer process, a report nobody had time to write. The point is proof that you can deliver inside this team's real constraints.

Days 61 to 90: Lead

Propose the bigger bet. Show the problem, the options, your recommendation and how you will measure it. Managers remember the person who arrived with a plan and then followed it.

Three habits that make it stick

Write each phase as outcomes with a measure.
Share a draft with your manager in week one and ask what they would change.
Review it every Friday for ten minutes and log what moved.

If you are interviewing

The same structure works as a six-slide presentation: the role, what you would learn, what you would contribute, how you would lead, how you would measure it, and the risks you see. Practice the tough questions out loud, especially "what if the plan changes?"

A ready-made kit

I put together the 30-60-90 Day Plan Kit: an editable Word template, five role examples (engineer, product manager, marketing, sales, new manager), an interview presentation script with twelve answer frames, a first-90-days tracker and a 12-page playbook. It is $12: https://logi2.gumroad.com/l/voskbb

If you want one practical template or guide each week, subscribe to the newsletter: https://aifounderweekly.substack.com

A 5-minute weekly habit that makes self-reviews and promotion cases easy

Natraj L — Sun, 04 Oct 2026 14:22:03 +0000

Every year the same thing happens. Review season arrives, your manager asks for a self-review, and you stare at a blank page trying to recall what you shipped ten months ago.

Managers rarely under-rate people out of bad faith. They forget, and what they forget they cannot reward. In a calibration meeting, the person with specific evidence beats the person with a vague impression almost every time.

The fix: a weekly log, not a better memory

1. Capture every Friday (5 minutes). Write what you shipped, who it helped, and any number you can attach: hours saved, tickets closed, latency cut, people unblocked. A rough number beats a missing one.

2. Group into themes. Two weeks before reviews are due, sort entries into three or four themes your company rewards: delivery, collaboration, technical depth, leadership.

3. Write outcomes, not tasks. Instead of "Worked on the onboarding flow", write "Rebuilt the onboarding flow, cutting setup time from 12 minutes to 4 and lifting week-one activation." Same work, far stronger line.

4. Make your manager's job easy. Send a one-page impact summary a week before calibration. The hardest part of promoting someone is finding specific evidence quickly, so hand it over.

If you want the templates ready-made

I packaged this as the Self-Review and Promotion Case Kit: a brag document, an annual self-review template with worked examples, a promotion case one-pager, and email scripts for asking for feedback or a level discussion. Four editable Word files plus a 9-page PDF playbook, $9: https://logi2.gumroad.com/l/hpnbq

Disclosure: I am the author of the kit. The habit works fine without it, so start the Friday log this week either way.

You Say Yes Before You Check If the Week Has Room. Here's the Fix I Built.

Natraj L — Sun, 27 Sep 2026 17:53:31 +0000

A new lead messages you: "Can you start Tuesday?" You say yes. It feels fine — you're not that busy. Three weeks later you're writing an awkward email to an existing client explaining why their deadline slipped, because nobody, including you, noticed that Tuesday didn't actually have room.

This is the single most common way freelancers and solo consultants get overbooked: not through bad planning, but through no planning — no number to check before saying yes. A calendar shows you what's scheduled. It doesn't tell you what's left.

The fix isn't more discipline. It's a number.

I built a workbook that replaces the guess with a number. You log the hours you've booked per client, per week, and nine linked tabs do the rest:

Setup — your real weekly available hours, target utilization %, and buffer time, set once
Clients & Rates — every active and pipeline client in one color-coded table
Weekly Capacity Log — 16 weeks pre-built, auto-summing booked hours per client
Capacity Dashboard — this week's snapshot plus an 8-week "overbooking radar" that flags a red week while you can still renegotiate it, not after you've already missed it
Client Deadline Tracker — days-remaining formulas that turn red under 3 days
Rate Reality Check — your true effective hourly rate after unbilled admin time, versus what you quote

No formulas to build. You type hours into the yellow input cells; everything else — the utilization percentage, the overbooking flags, the revenue-at-risk estimate — calculates itself.

Why this matters more than another to-do app

Task managers tell you what to do. They don't tell you whether you have the hours to do it. The gap between those two things is exactly where overcommitment lives, and it's invisible until the week you're already in it.

The workbook also comes with a 9-page companion PDF guide, including three ready-to-send scripts for the exact moments this is built for: a new lead wants to start now, an existing client adds scope mid-project, and the week you realize you're already overbooked.

Who this is for

Freelance consultants, designers, developers, and writers juggling three or more active clients — anyone who has said yes to a project and then spent the next hour doing capacity math in their head instead of in a tool built for it.

It works in Excel natively, and imports cleanly into Google Sheets (File > Import) with every formula and conditional-format rule intact.

If you've ever found out you were overbooked the same week it happened, this is built to make sure that doesn't happen again.

Get it here:

Gumroad ($39): https://logi2.gumroad.com/l/cvgbq
Etsy ($27): https://www.etsy.com/listing/4583655465/freelancer-capacity-planner-spreadsheet

You remember a third of your 1:1s - so I built the system I should have had

Natraj L — Sun, 20 Sep 2026 17:19:31 +0000

A manager with six reports holds roughly 150 one-to-ones a year. Ask most of us what was said in any given one three weeks later and you get an impression, not a record.

Then review season arrives and a year of someone's work gets reconstructed from the last six weeks anyone can actually recall. That is how recency bias ends up inside a salary decision.

The four ways 1:1s decay

I have watched all of these happen on my own calendar:

The status meeting. You already know what they are working on - it is in the tracker. If the 1:1 is a spoken version of the board, you are spending your most valuable half hour reading something you could have read in ninety seconds.
The meeting with no memory. Good conversation, real problems, honest answers, and none of it written down. Three weeks later you cannot remember whether you promised to talk to the payments lead or they did.
The meeting that keeps moving. Each cancellation is individually rational. The pattern is not, and your report is counting even if you are not.
The meeting where nothing hard is ever said. Warm, pleasant, and slightly useless - because the hard feedback got saved for the review cycle, where it lands as a surprise.

What I built

A 9-tab spreadsheet and a 14-page playbook. Nothing clever, just a place for the record to live:

1-1 Log - 200 rows: their topics, your topics, feedback owed in both directions, what was decided, who owes what, and a 1-5 energy read that turns red on its own.
Action Items - every promise made in a 1:1, with an owner and a self-calculating "days late" column. Filter by Owner = Me once a week and be honest about what you see.
Growth Evidence - one line per observed thing, tagged to a competency and a scope. Promotion cases are not written in November; they are collected all year.
Review Prep and Team Health - meetings held, last met, days since, open actions both ways, evidence count, average energy. All calculated, nothing typed.
Question Bank - 60 questions sorted by the situation you are actually in: a new report, someone stuck, low energy, hard feedback, after a bad incident, someone clearly job-hunting.

The whole thing is three habits: ten minutes before each 1:1, five minutes after, ten minutes on Friday.

The one rule that makes it work

Write the log inside the meeting or within ten minutes of it. Not at the end of the day. Not on Friday. A 1:1 written up on Friday is a 1:1 you half-remember, which is where you started.

If you take notes in the room, say so once out loud: "I write these down so I don't drop things you asked me for." Nobody has ever objected to that sentence.

Where to get it

It works in Excel, Google Sheets, LibreOffice and Numbers. No macros, no add-ins, no account.

Gumroad: https://logi2.gumroad.com/l/uurnfx (\$9)
Etsy: https://www.etsy.com/listing/4579037146/11-meeting-template-for-managers-one-on

I write up one of these tools a week - workbooks, playbooks, small things that solve one specific problem. If that is useful, the newsletter is free: https://aifounderweekly.substack.com

Also published on Medium.

Your MRR Is a Guess Until You Check It. Here's How I Fixed That.

Natraj L — Sun, 13 Sep 2026 13:57:18 +0000

I once had an investor call in twenty minutes and no idea what my actual MRR was. I had a Stripe dashboard open in one tab, a spreadsheet from three weeks ago in another, and a mental note that "a couple people churned recently." I was reconstructing my own business's numbers live, on a call, from memory.

Most solo SaaS founders do exactly this. Revenue lives in Stripe, expenses live in a bank statement or a different spreadsheet, and churn lives in your memory of who emailed to cancel. None of it is linked, so every investor update or founder check-in turns into twenty minutes of manual math before you can even talk about the number.

So I built a workbook that does the math for you. It's a 7-tab Google Sheets / Excel file — Subscribers, Churn Log, Expenses, Plans & Pricing, Settings, Start Here, and a Dashboard — where the only tabs you ever touch are Subscribers, Churn Log, and Expenses. Everything else — MRR, ARR, churn rate, ARPU, CAC, estimated LTV, LTV:CAC ratio, and cash runway in months — calculates itself from formulas the moment you add a row.

It also watches itself. Two built-in alerts flag it automatically if your churn rate climbs above 5% or your LTV:CAC ratio drops below your own target, so you're not relying on noticing a bad trend by eye. A 7-page walkthrough guide explains exactly how each metric is calculated, so the numbers aren't a black box either.

The full workbook plus the guide is $47 on Gumroad, $27 on Etsy, and $39 on Payhip. Set it up once, in under ten minutes, and you'll never do investor-call math from memory again.

Originally published on Medium.

Scope Creep Isn't a Client Problem. It's a Process Problem.

Natraj L — Sun, 06 Sep 2026 18:47:17 +0000

I once spent forty minutes on a call talking a client out of "just one more small tweak" that was, by my own estimate, about six hours of unbilled design work. I didn't lose that fight because I lacked a contract. I lost it because my contract didn't give me anything to say in the moment. It just sat in a drawer, unread since the day it was signed.

Most freelancer contracts are like this: solid on liability and payment terms, silent on the specific moments that actually cost you money. Scope creep doesn't show up as a legal dispute. It shows up as "can you just also..." on a Tuesday afternoon, and there's no clause telling you what to say back. So you say yes, because saying no feels confrontational and you don't have a script ready. Multiply that across every client, every quarter, and it's not a contract problem. It's a process problem: nobody wrote down what happens when a request falls outside the original scope.

So I built The Client Onboarding & Contract Playbook, a 12-page plain-English guide for freelancers and solo consultants who already have some form of contract but are missing the parts that actually get used under pressure. It covers a 5-step onboarding sequence that catches scope questions before a project starts, 8 contract clauses explained in plain English instead of legalese you'd have to translate for a client anyway, a 10-question client intake questionnaire, and copy-paste scripts for the three conversations freelancers dread most: pushing back on scope creep, following up on a late invoice at 7/20/30+ days, and firing a client who's no longer worth the account.

This isn't a contract generator. If you need something legally binding for your jurisdiction, talk to a lawyer. This is for the freelancer who already has a contract and still finds themselves agreeing to extra work they never signed up for, because in the moment, silence is easier than confrontation. The scripts exist so you have language ready before the moment happens, not while you're on the call scrambling to sound reasonable.

The full guide is $17 on Gumroad and $12 on Etsy. If you want more of this — actual freelance-business friction, not another productivity platitude — I write about it weekly at aifounderweekly.substack.com.

Originally published on Medium.

I stopped hiding my career pivot on my resume and started getting callbacks

Natraj L — Sun, 30 Aug 2026 18:00:32 +0000

For three months I applied to "AI product manager" and "AI program manager" roles with a resume that listed my actual job titles and nothing else. Silence, mostly, with a few generic rejections. The pattern only became obvious once I compared notes with someone on the other side of the table: hiring managers weren't doubting my competence. They were doubting my seriousness. A resume that doesn't name the pivot reads like someone chasing a trend, not making a real move.

The fix wasn't a better summary line. It was naming the pivot directly, in the first sentence, and then backing it with proof instead of adjectives — one real project, documented, not implied. "Led cross-functional teams" became something closer to "Built and shipped an internal AI workflow tool that cut manual review time in half," even when that project started as a weekend build rather than an official assignment.

The second unlock was the LinkedIn "About" section, which almost nobody rewrites when they pivot. Mine still described my old role for months after I'd started applying elsewhere — a small thing, but it's usually the first thing a recruiter checks after the resume, and it was quietly contradicting everything I was trying to say.

Once both were aligned — resume naming the pivot with proof, LinkedIn backing it up — the response rate changed. Not overnight, and not because of a trick. Just because the story finally made sense on both documents at the same time.

I turned the templates, the project framework, and the follow-up email sequence I used into a kit for anyone doing the same pivot — 1-page and 2-page ATS resumes, a cover letter structured paragraph by paragraph, the LinkedIn framework, three follow-up email templates, and a step-by-step guide for building your one AI project if you don't have one yet: The AI Career Pivot Resume Kit.

If you're mid-pivot right now and the silence is getting to you, it's usually not about your competence. It's about whether your documents make the pivot legible.

Originally published on Medium.

My Freelance Business Lived in Four Apps That Never Talked to Each Other

Natraj L — Sun, 30 Aug 2026 13:29:27 +0000

Every Monday morning, before I could do any actual client work, I'd spend 20 to 30 minutes just reconstructing where things stood. Not because I'm disorganized. Because my freelance business was scattered across four tools that never talked to each other.

Leads lived in a spreadsheet. Deliverables lived in a task app. Invoices lived in a doc. And the actual status of each relationship lived in my head. That last part is the dangerous one - memory doesn't scale past 3 or 4 active clients.

So I rebuilt it inside Notion as one connected workspace: six linked databases that reference each other automatically (Command Center Dashboard, Client CRM, Project and Task Tracker, Invoice and Revenue Log, Client Onboarding Checklist, Weekly Business Review). Open a client and you see their open tasks and outstanding invoices without re-typing anything. Runs on Notion's free plan, no formulas required.

I packaged the build as the Notion Freelance Business Hub: a 15-page visual walkthrough plus a copy-paste setup guide with the exact property, relation, and rollup order. Setup takes 60-90 minutes, or spread it across a week.

Get it on Gumroad or Etsy.

If you want more build-in-public breakdowns like this, I write a free weekly newsletter at aifounderweekly.substack.com - one click to subscribe, no spam.

Originally published on Medium.

I almost accepted the lower offer because the bigger number was harder to compare

Natraj L — Sun, 23 Aug 2026 14:26:10 +0000

Two offers landed in the same week. One was a straight $105K in a city I already lived in. The other was $95K, fully remote, with equity, a better 401(k) match, and a relocation stipend I wasn't planning to use. My gut said take the $105K -- it was just a bigger number, and bigger numbers feel safer when you're deciding in a hurry.

I almost did it. Then I actually sat down and did the math instead of eyeballing it, and the remote offer won by a wide margin once I accounted for the equity vesting on a normal schedule, the 401(k) match difference over a year, and what $95K actually buys in a place with a lower cost of living than where the $105K job was based. Base salary alone hid more than it revealed.

That's the trap with comparing job offers: the number on top is the easiest thing to compare and the least reliable one. A $95K offer with real equity and a strong match can beat a flat $105K offer once you do the math. A remote offer tied to an expensive city's pay band can be worth less than a smaller number somewhere cheaper. Most people either eyeball it under time pressure, or open a blank spreadsheet the night before they have to respond and try to rebuild all of this from scratch.

So I built the workbook I wish I'd had that week: The Job Offer Decision Kit.

It's a 6-tab spreadsheet plus a companion PDF guide that turns base salary, bonus, signing bonus, equity/RSU grant, 401(k) match, benefits, and PTO into one clear Year-1 number for up to 4 offers side by side. It adjusts that number for cost of living, so a bigger salary in an expensive city gets compared honestly against a smaller one somewhere cheaper. It scores the non-money factors -- growth, culture, stability, flexibility -- by what actually matters to you, not a generic checklist. And it sets your target ask and your walk-away number before you get on the phone, not during it, with 6 ready-to-send negotiation scripts for the conversations that actually come up.

You don't need to know spreadsheet formulas to use it. Every cell you're meant to fill in is color-coded with an example value already in it; everything else calculates automatically. It works in Excel, LibreOffice, and Google Sheets, and setup takes about 10-15 minutes for one offer.

If you've got an offer sitting in your inbox right now, or you're expecting one soon, you can get it here: https://logi2.gumroad.com/l/ukkvqm

Originally published on Medium.

I Stopped Trusting Green CI on AI-Generated Pull Requests

Natraj L — Sun, 16 Aug 2026 15:06:53 +0000

Three weeks ago I merged a PR that Claude Code wrote for a side project.

The diff was clean, the tests passed, CI was green. I skimmed it, approved it, moved on. Two weeks later I found out the authorization check on a delete endpoint had quietly disappeared during a refactor earlier in that same session. Nobody exploited it. I just got lucky.

That's when it clicked that AI-generated code doesn't fail the way human-written code fails. A human engineer forgets an edge case or doesn't know a library has a built-in for something. An AI coding agent writes a plausible-looking auth check, a plausible-looking test, a plausible-looking API call, and "plausible" is exactly the failure mode that slips past a reviewer skimming a diff at 6pm. None of it shows up as a red squiggly line. It just looks finished.

So I went back through a stack of AI-generated PRs, mine and a few teammates shared with me, looking for a pattern instead of individual bugs. Six showed up over and over: hardcoded secrets dropped in for a "working example," authorization checks that check you're logged in but not that you're allowed to touch this specific resource, tests that mock the exact function they're supposed to be testing (so they pass no matter what the implementation does), API calls that look right but don't match the actually-installed library version, edge cases that fail silently instead of throwing, and scope creep where the agent "improves" an adjacent file nobody asked it to touch.

I built a toolkit around those six patterns: what each one looks like in a real diff, the exact review question that catches it, a 45-item checklist organized by risk category, a 1-5 risk scoring rubric so you know how deep to review before you start, and a companion spreadsheet that logs every PR you review so the pattern becomes visible over time instead of staying anecdotal, which AI tool in your stack actually produces the most rework, not a guess.

It's for solo developers and small teams (2-15 people) who've adopted Claude Code, Cursor, or GitHub Copilot for a meaningful share of their code and are currently reviewing AI-generated PRs the same way they'd review a junior engineer's: reading the diff, trusting green CI, moving on. If that's you, the 5-minute triage checklist alone is worth running on your last three merged PRs before you read anything else. Most people find something on the first pass.

If you want the full toolkit, a 16-page PDF plus the PR risk tracker workbook, it's here: https://logi2.gumroad.com/l/lpdebo

The thing I keep coming back to is that "the code looks finished" and "the code has actually been checked" are two different claims, and AI coding agents are very good at making the first one true regardless of the second. Worth remembering the next time a diff looks a little too clean.

Originally published on Medium.

I Built a Google Sheets Calculator That Tells Tech Workers Exactly When They Can Quit Their Job

Natraj L — Fri, 14 Aug 2026 02:00:12 +0000

Six months ago, I had a Notion doc titled "The Plan." It had my side hustle revenue projections, my savings runway, my client pipeline. I thought it was thorough. Then my manager scheduled a surprise 1:1 and I immediately started refreshing my bank balance on my phone under the desk.

The spreadsheet wasn't the problem. The spreadsheet was missing something: a single, clear answer to the question I was actually asking. Not "how much do I make freelancing?" but "can I afford to quit right now?"

So I built something that gives you that answer.

What I built

The Side Hustle Full-Time Decision Planner is a Google Sheets workbook for tech workers who are seriously considering going independent. Not daydreaming — actually close to pulling the trigger, or actively trying to figure out when they can.

It has four connected calculators:

Financial Runway Calculator. Enter your monthly expenses, current savings, and minimum acceptable income. It tells you exactly how many months you can survive without a single client.

Client Income Projector. Model what your income looks like at 3, 6, and 12 months post-quit: conservative (1 client), realistic (3 clients), and aggressive (5+ clients).

Risk Score Dashboard. Scores your health insurance plan, estimated tax setup, and emergency fund readiness.

Go/Wait Decision Engine. Produces one output: a recommended quit date, or the specific milestone you need to hit first.

Who this is for

Software engineers, PMs, designers, and data professionals earning $80K–$200K who have a side hustle generating real income and are trying to make the leap with math rather than motivation.

Where to get it

Available on Gumroad at $17: https://logi2.gumroad.com/l/sidehustleplanner

Also on Etsy and Payhip. Works in Google Sheets (free) and Excel. Includes a PDF companion guide.

Originally published on Medium