DEV Community: owen zhang

Intercom vs Zendesk: what I actually learned after using both

owen zhang — Wed, 10 Jun 2026 15:06:08 +0000

When we launched our SaaS, we picked Zendesk. Not because we researched it — because everyone we knew used Zendesk. It was the default.

Eighteen months later we evaluated Intercom. Not because Zendesk was broken — because our support workflow was getting expensive in ways that didn't show up in the license cost.

Here's what I wish someone had told me before we spent three weeks migrating.

Zendesk is a ticket machine. Intercom is a messaging platform.

This distinction sounds like marketing copy but it actually determines whether the software fits your team's mental model.

Zendesk is built around tickets. A customer has a problem → ticket opens → ticket closes. The queue is clean. The audit trail is clean. Your support manager can run SLA reports without touching a spreadsheet. If you have a dedicated support team, this flow feels natural.

Intercom builds around conversations. Every exchange lives in a persistent thread. This sounds more human — and it is, for small volumes. But at scale, we started seeing threads revived from two months ago, reps accidentally pinging customers who'd long since moved on. The UI doesn't make it obvious which threads are dormant.

Pricing transparency: neither vendor wins here

Zendesk's entry price looks manageable until you hit the features list. Custom CSAT surveys, advanced reporting, and SLA management all live behind the next tier up. We hit that ceiling faster than expected.

Intercom charges per seat AND per AI resolution (their Fin chatbot). When we had a chatty month, we got a surprise invoice. No soft warning, no spend cap at that plan level.

The lesson: always run both through a full pricing simulation before committing. Take your actual monthly conversation volume, plug it into each vendor's pricing page, and model what happens at 2x growth.

Integration depth

Zendesk has a giant marketplace. Most major tools have connectors. But many integrations are shallow — they notify you that something happened without letting you act on it from inside Zendesk.

Intercom's Messenger embeds directly in your product UI, which means customers get help in context rather than being routed to a separate portal. That's a genuinely better user experience. The trade-off: Intercom becomes load-bearing infrastructure. When we considered switching away, the migration estimate was three weeks. You don't just "move off" Intercom — you also rebuild onboarding flows, in-app banners, and proactive messaging sequences.

The honest use-case split

Zendesk fits if:

You have a dedicated support team separate from engineering
You're handling 100+ tickets per day
Enterprise SLAs matter to your customers (audit trail, escalation rules)

Intercom fits if:

Support and product teams share responsibility for the customer experience
You want in-app onboarding and messaging bundled with support
You're early-stage and per-seat costs are still predictable

One thing that tripped us up

Neither tool makes it easy to trial with real production data. Both offer sandboxes, but the sandbox doesn't replicate your actual conversation volume or team workflow. You won't feel the real difference until you're two months in with real customers.

If you're making this call right now, I put together a more detailed Intercom vs Zendesk breakdown with current pricing, side-by-side features, and a decision framework — might save you some of the discovery time.

Which one are you on? And if you've switched, what actually pushed you over the edge?

How I finally stopped dreading payroll day (what changed after switching tools)

owen zhang — Tue, 09 Jun 2026 09:31:08 +0000

For the first three years of running a small dev shop (just me and four contractors), I handled payroll manually. Spreadsheet, bank transfer, a prayer that I hadn't transposed anyone's hours. It worked — until it didn't.

The breaking point was a mid-year audit flag on contractor vs. employee classification. Nothing serious, but it exposed how much I'd been winging the compliance side. That was when I actually sat down and evaluated payroll tools properly instead of just Googling "cheap payroll software" and picking whatever had a free trial.

Here's what I learned after testing six different platforms over about four months.

The real cost isn't the monthly fee

Every payroll tool has a base fee ($6–$40/month) plus a per-employee or per-contractor charge. For a small team under ten people, the base fee barely matters — the per-run or per-person charge is what adds up.

What caught me off guard: some tools charge per payroll run, not per month. If you do biweekly payroll for five people, that's 26 runs/year. A platform that looks cheap at $5/run turns into $130 just in run fees before you've paid for a single employee.

Gusto's pricing is $40/month + $6/person. For five people it's $70/month — that felt steep at first. But I was comparing it to tools that charged less monthly and more per-run, and when I did the math, Gusto came out similar or cheaper for my actual usage pattern.

Contractor-only is a different use case

I have a mix: two actual employees (me + one full-time), and three contractors. This caused me real headaches because most tools price employees and contractors very differently.

Gusto handles both in the same interface, with 1099s included in most plans
Rippling is overkill for under 10 people but incredibly clean for mixed teams — the HR module is baked in rather than bolted on
Wave Payroll is cheap but contractors are almost an afterthought in the UX

If you're fully contractor-based (common in dev shops), you might be better served by a tool that focuses on 1099 management rather than trying to use a full payroll system.

What "automated tax filing" actually means

Every payroll tool claims to handle taxes. What varies is which taxes and how much they do automatically vs. what you still need to touch.

The basics (federal + most state payroll taxes) — nearly all tools handle this. Where they diverge:

Multi-state employees: if even one employee is in a different state, some tools charge extra or require manual setup
New hire reporting: legally required, often forgotten — Gusto and Rippling both automate this
Year-end W-2/1099 filings: included in most mid-tier plans, but sometimes a paid add-on on entry-level plans

Read the fine print on what "full-service payroll" means in each plan tier before assuming it handles your specific situation.

The three questions I'd ask before picking anything

Does it handle my specific mix? (employees only, contractors only, or both)
What's the all-in cost at my exact headcount? (base + per person + any per-run fees)
Does it automate my state specifically? (a few states have unusual requirements that some tools don't fully cover)

I ended up on Gusto — not because it "won" some feature comparison, but because it handled my 2-employee + 3-contractor mix cleanly, the tax filing actually worked without me having to do anything after setup, and the interface doesn't feel like it was designed in 2009.

If you want a fuller breakdown across more tools — including Rippling, Deel (which I looked at for the contractor side), and a few budget options — I went deep on this at HRPayPick's payroll software guide. Might save you the four months of trial-and-error.

One thing I wish someone had told me earlier: don't pick payroll software based on what's popular for 500-person companies. The priorities are completely different when you're under 10 people and the "HR team" is also the founder.

Picking a help desk for a tiny SaaS team without overpaying

owen zhang — Tue, 09 Jun 2026 07:12:39 +0000

When our side project picked up its first paying users, support went from a couple of emails a week to a Slack channel full of "is this a bug or am I being dumb" messages. We needed an actual help desk. Picking one turned into a rabbit hole, so here is the short version I wish someone had handed me.

The pricing trap nobody mentions

Most help desks advertise a low per-agent price, then the feature you actually need lives two tiers up. Zendesk starts around $19/agent/month, but the AI add-on is roughly $50/agent/month on top, and the reporting most teams want sits on a higher Suite tier. Freshdesk does something similar once you move from the basic support desk to the omnichannel plan with chat and phone.

So compare the realistic 12-month config, not the entry price.

What actually matters at 1 to 5 people

For a tiny team, three things beat everything else:

Time to first reply, not feature count. You want to be answering tickets the same afternoon you sign up.
A shared-inbox feel over a heavy ticket-system feel. Early users hate being a ticket number, and your team hates context-switching into a bloated tool.
A free or near-free tier so you are not paying $300/month before you have product-market fit.

The shortlist that kept coming up

Freshdesk Free: up to 10 agents at $0. Honestly hard to beat as a starting point.
Help Scout (~$22/user/month): reads like email, not tickets. Good fit for a small SaaS that cares about tone.
Zoho Desk (~$14/agent/month): the value pick, especially if you already use other Zoho tools.
Zendesk: powerful and scales far, but you feel the complexity and cost early. Worth it later, overkill at 3 people.

What I would actually do

Start on Freshdesk Free or Zoho Desk. Keep your FAQ and docs in the same tool so you deflect the repeat questions instead of retyping them. Only graduate to Zendesk when you have real SLAs, multiple queues, and the volume to justify the setup time.

If it helps, I put together a fuller comparison with current pricing and the gap between G2 scores and real-world complaints here: https://commsadvisor.com/freshdesk-vs-zendesk/

Curious what everyone else landed on for early-stage support. Did you start with a real help desk, or just live in a shared Gmail until it broke?