This post in 30 seconds.

• The real choice isn't two-way. For your phone line it's three: an in-house team, a human BPO, or an AI phone agent. Per resolved call is what decides it.
• A loaded in-house rep runs about $4,000/month ($90,000 to $110,000/year fully burdened). A human BPO runs $1.50 to $3.50 per call. An AI phone agent runs about $0.42 to $0.91 per resolved call.
• Written for founders, COOs, and Heads of CX at $10M-$100M Shopify brands running 3-12 reps, a paid helpdesk, and a phone number that's still on the site.

The decision almost never starts as a strategy doc. A rep gives notice, or goes on leave, or the support line on the P&L creeps up another point, and someone asks the real question: do we hire #5, or do we send this out?

That's where most brands frame it as a coin flip. In-house or outsource. Build it or buy it. But for the channel that actually leaks revenue, the phone, that binary is the wrong frame in 2026. There's a third option now, and on the routine calls it's cheaper per resolution than either of the other two. This is the math, with real numbers, so you can decide which one (or which mix) your store actually needs.

If support is getting too expensive to handle in-house and the after-hours calls are the part that keeps slipping, book a 30-min call and we'll run your numbers with you. We've launched AI phone agents for 50+ Shopify brands sitting in exactly this spot.

In this post:

• Why it's a three-way decision, not two
• What each option actually costs per resolved call
• The AI phone agent: the third path
• How to decide (and the stack that works)
• Frequently asked questions

The decision isn't in-house vs outsource. It's three ways.

Almost every comparison you'll read treats this as two boxes: an internal team you manage, or a vendor you pay to manage it for you. Both are real. Neither is the whole picture anymore.

For the phone channel specifically, you now have three live options, not two: an in-house team, an outsourced human BPO, and an AI phone agent that takes the routine calls and hands the rest to a person. The reason this matters is that phone is the channel where the three options diverge hardest on cost.

Email and chat are forgiving. A ticket can sit for an hour and nobody churns. The phone isn't like that. Businesses answer only 37.8% of inbound calls, with the rest going to voicemail or nothing at all, according to AmbsCallCenter. And once a caller can't reach a person, 85% never call back and 62% go buy from a competitor (PCN).

So the phone is the channel with the most expensive failure mode, the most after-hours leak, and the highest share of the same questions over and over. WISMO calls ("where's my order"), returns, "is this in stock." That mix is exactly what changes the calculus on who, or what, should pick it up. Most ecommerce customer service advice ignores this and treats every channel the same, which is why so many brands end up outsourcing all of it to one vendor when a split would cost far less.

What each option actually costs per resolved call

Here's the part the other comparison posts skip. They'll tell you in-house is "higher fixed cost" and outsourcing is "lower variable cost" and leave it there. No numbers. So let's put real numbers on all three.

The costs below come from real Ringly billing across 50+ Shopify brands, plus the standard loaded-cost math for human reps and the published 2026 BPO pricing guides. Where a number is a range, I'll say so.

An in-house rep is the most expensive way to answer a routine call, and it's not close. A US customer service rep earns about $40,500 a year in base salary in 2026 (WowRemoteTeams). But the loaded cost is what hits your P&L. Add benefits, payroll tax, software, office overhead, training, and QA, and one agent runs $90,000 to $110,000 a year, per ClearSource. The fully burdened cost lands at 1.3 to 1.4 times base salary as a rule of thumb (HiBob). Call it roughly $4,000/month loaded, which works out to about $2.70 per call.

Then there's the part nobody quotes you on. Average rep tenure is 13 to 15 months, and replacing one costs $10,000 to $20,000 in recruiting, training, and lost productivity (4 Corner Resources). So your loaded cost isn't even stable. It resets every time someone burns out and quits.

A human BPO is cheaper per hour and gives you human coverage you don't have to hire. Offshore voice support runs $7 to $16 an hour, Latin America $12 to $19, and onshore US specialists up to $65, per Crescendo's 2026 pricing guide. Per call, that lands around $1.50 to $3.50. But the sticker rate hides a lot: onboarding fees of $500 to $2,000 per agent, a 3-to-6-week ramp, weekend and overnight premiums of 15 to 50%, contract minimums, and add-on charges for analytics and call recordings (Text.com).

An AI phone agent is the cheapest per resolution on routine volume. Across 50+ brands, the AI resolves 73% of inbound calls on its own at roughly $0.42 per resolved call, with zero ramp and 24/7 coverage at no premium. WashCo, a Shopify brand we launched, recovered $22,664 in its first 7 days on the phone.

Here's all three on one table.

The takeaway isn't "fire the humans." It's that the routine 70% of your phone calls (order status, returns, product questions) shouldn't be priced at $2.70 a call when the same answer costs $0.42. The hard 30%, the genuinely complex or emotional calls, are where a human's time is worth paying for.

In-house support: when owning it pays off

Owning your support team isn't the wrong answer. For a specific kind of call, it's the only good one.

An in-house team is worth the loaded cost when the call requires judgment, brand voice, or sensitive data, not when it's the fifth WISMO of the hour. If your product is genuinely complex, if a wrong answer carries real liability, or if your customers expect a relationship rather than a resolution, a person who knows your brand inside out earns their seat.

Best for: complex or high-trust calls, regulated products, and brands where the support conversation is part of the customer relationship.

What works

• Brand voice and product depth. Your own reps live in your catalog and talk like your brand, not a script handed over last week.
• Data control. Customer information never leaves your walls, which matters if compliance is tight.
• Accountability. When something breaks, the person who owns it is on your payroll, not three time zones away.

What doesn't

• The loaded cost scales linearly. Every jump in call volume is another $90K-$110K/year hire (ClearSource).
• Tenure is short. At 13-15 months average, you're re-hiring and re-training constantly (4 Corner Resources).
• After-hours is a gap. Staffing a night shift means paying reps to sit idle through the quiet hours, because the volume isn't steady. 24/7 phone coverage almost never pencils out with humans alone.
• Burnout. Reps hired to solve real problems quit fast when the job is answering the same five questions all day.

Outsourced human BPO: flexible coverage, real trade-offs

The pull toward a BPO for Shopify is usually coverage, not cost. You need humans on the phone at hours you can't staff, or a buffer for the seasonal spike, and you don't want to run the hiring yourself.

A human BPO buys you flexible human coverage fast, but you trade away brand voice, ramp speed, and a chunk of the savings to hidden fees. It's the right move when you genuinely need more people, not more efficiency, and you need them across hours your own team can't cover.

Best for: bursty human coverage, multilingual support, and fast headcount flex without internal recruiting.

What works

• 24/7 human coverage without you building a night shift.
• Fast scaling for launch weeks and holiday peaks.
• No recruiting burden. The vendor hires, trains, and manages.
• Variable cost that flexes with volume, at least in theory.

What doesn't

• Brand voice drifts. Agents juggle multiple clients, so you're handing over scripts and FAQs up front and hoping they stick. Merchants on the Shopify community consistently flag this as the catch.
• Ramp is slow. Plan on 3 to 6 weeks before agents are productive, plus $500-$2,000 per agent to onboard (Text.com).
• Hidden fees eat the savings. Weekend and overnight premiums, contract minimums, early-reduction penalties, add-on charges.
• Data exposure. A third party now touches your customer data.

There's also a quality risk that doesn't show up until later. Undertrained agents resolve the call but leave the customer slightly worse off, and that shows up downstream as churn. If you go this route, the brands that make it work are the ones who treat the BPO as a tier-2 escalation layer, not the front line for every call.

The AI phone agent: the third option for the routine 70%

This is the path that didn't exist a few years ago, and it's the one the binary comparisons miss.

Ringly is AI phone support for Shopify. The pitch isn't "stop hiring people." It's "stop paying $2.70 a call for the question you've answered ten thousand times." The AI takes the routine inbound calls so your team can focus on the work that actually moves revenue.

The AI answers inbound calls 24/7. It finds orders in your Shopify store, processes returns and exchanges, answers product questions from your knowledge base, and rescues abandoned carts. Calls that need a person escalate cleanly through smart call transfer to Gorgias, Zendesk, or whatever helpdesk you already run. You keep your number, your stack, and control over what escalates.

The AI is the right tool for the routine 70%, not the hard 30%, and being honest about that line is the whole point. Order status, returns, "do you have this in my size", checking order status: the AI handles those end to end. A grieving customer, a genuinely complex edge case, a high-value account that wants the founder: those still go to a human. The AI just makes sure your humans aren't buried under WISMO before they get to them.

"My customers also feel like it's a normal person. They feel like they can communicate if they have questions."
Claudia Droge, TechCraft Studio

The most repeated thing customers say after a call is "you don't sound like AI," which is the bar that matters. If it sounds like a phone tree, customers hang up and the whole point is lost.

Pricing. Grow is $349/month (1,000 minutes, ~500 calls). Pro is $799/month (2,500 minutes, ~1,250 calls) with a 14-day free trial. Enterprise is by call only, for $10M-$100M brands with 3-12 reps. Live in under an hour. And it's backed by a 65% resolution guarantee: if the AI resolves under 65% of your calls in 90 days, we refund the last 3 months.

How to decide (and the stack that works)

You don't actually have to pick one. The best support operation at a $10M-$100M Shopify brand isn't a single model, it's a stack split by channel and by tier.

Choose an in-house team if your calls are complex, your product is regulated or high-trust, or the support conversation is part of how customers feel about your brand. Keep them on the hard 30%, not the WISMO queue.

Choose a human BPO if you need flexible human coverage across hours or languages you can't staff, and you're willing to manage the ramp and the brand-voice gap. Use them as a tier-2 escalation layer, not the front door.

Choose an AI phone agent if the bulk of your phone volume is routine and repeatable, after-hours is leaking, and the per-call cost of a human for those calls no longer makes sense. The pricing starts at a flat $349/month, which reframes the whole math against a $4,000/month hire.

The stack that actually works for most brands in this band looks like this. AI on the routine inbound phone calls. Helpdesk AI on tier-1 email and chat. A small in-house team or BPO on tier-2 escalations and the emotional calls. You stop paying premium human rates for the question a machine can answer in 30 seconds, and you free your people for the calls that actually need them.

Want to see which split makes sense for your store? Book a 30-min call and we'll do the math on your real call volume, live.

Frequently asked questions

Is outsourcing customer support cheaper than in-house? On a per-hour basis, usually yes: offshore BPO runs $7-$16/hour versus a loaded in-house rep at roughly $90K-$110K/year. But hidden fees (ramp, weekend premiums, minimums) and brand-voice loss narrow the gap, and for routine phone volume an AI phone agent beats both per resolved call.

How much does an in-house Shopify CS rep really cost? Base salary is about $40,500/year in 2026, but the fully loaded cost (benefits, tax, software, training, QA) runs $90,000 to $110,000 per agent. Add the $10K-$20K it costs to replace one every 13-15 months and the real number is higher than the salary line suggests.

When should a Shopify brand outsource support? When call volume outgrows what your team can handle and the next move is either hiring rep #5 or sending it out. For most $10M+ brands the smartest first step is sending the routine phone calls to an AI agent and reserving humans (in-house or BPO) for the complex ones.

Can an AI phone agent replace my CS team? No, and it shouldn't. It handles the routine 70% of phone calls (order status, returns, product questions) and escalates the rest to your team. The goal is to stop your reps from drowning in repetitive calls, not to remove them.

Will an AI phone agent sound like a robot? The most common thing Ringly customers report hearing from callers is "you don't sound like AI." Voice quality is the line that decides whether customers stay on the call or hang up, so it's the thing we obsess over.

Does AI phone support work with Gorgias or my existing helpdesk? Yes. Calls that need a human escalate cleanly to Gorgias, Richpanel, Reamaze, or whatever helpdesk you already run. You keep your number and your stack and control exactly what gets transferred.

What's the best support stack for a $10M+ Shopify brand? AI on routine inbound phone calls, helpdesk AI on tier-1 email and chat, and a small in-house team or BPO on tier-2 escalations and emotional calls. It puts the cheapest tool on the highest-volume work and saves human time for where it counts.

Talk to us

If you run a $10M-$100M Shopify brand and the phone is where calls leak after-hours, a 30-min call is the fastest way to see what the routine volume is actually costing you. We'll look at your real numbers and tell you honestly which split (in-house, BPO, AI, or a mix) fits your store.

The 3-layer guarantee.

1. Live in 14 days or it's free until launched.
2. 65% resolution in 90 days or we refund the last 3 months of subscription fees.
3. We keep working free until we hit it.

Ruben (Ringly co-founder) takes these calls personally.

Book a 30-min call →