Build vs buy

You could build this on Lob.
The printing was never the hard part.

We should say this first: Sincerely runs on Lob. Your postcards and letters are printed and mailed by the same vendor you were considering calling directly, and the markup is disclosed on the pricing page.

So the honest question is not whether Lob is good. It is. The question is what you would still be holding after the API returns a tracking number.

Start freeSchedule a call

The channel, by layer

Incrementality and reportingSincerely
Rep follow-up, timed to deliverySincerely
Orchestration, cadence, guardrailsSincerely
Approval, versioning, provenanceSincerely
Research and personalized writingSincerely
Audience, signals, deliverable addressesSincerely
Print, postage, delivery trackingLob & co

A print API is the bottom band. Everything above it is the part a revenue team actually experiences, and the part you would be building.

The category

Nobody asks why you don’t
run campaigns out of SendGrid.

Every channel your team already runs has two layers: a primitive that moves the message, and a platform that decides who gets it, what it says, and whether it worked. You have bought the second layer four times already without anyone proposing you wire up the first one yourself.

SendGridOutreach, Salesloftemail
TwilioAttentive, Klaviyosms
S3 and EC2Vercel, Netlifyhosting
StripeChargebee, Maxiobilling
LobSincerelymail

Lob is excellent infrastructure, and infrastructure is supposed to be boring and interchangeable. The channel is the product.

The work

Printing is one stage of eight.

Here is the whole job, and who does each part if you integrate a print API yourself. One of these is an engineering problem. The rest are not, which is why they tend to be discovered in month three.

Pick the accountsTurn CRM signals into an audience worth the postage, and keep it fresh as deals move.you build it
Find a deliverable addressLob will verify an address you already have. Getting one for a contact you know only as a work email, on a team that no longer sits in the HQ on the account record, is a different problem and an ongoing data-vendor relationship.you build it
Research the accountThe round, the roles they opened, the thing they shipped. Gathered per account, kept current, and attached to the claim it supports.you build it
Write every pieceIndividually, from that research, in a voice that survives a CEO reading it, at a volume no one is writing by hand.you build it
Approve before it printsA human queue, versioned edits, and weak claims flagged rather than quietly shipped.you build it
Print, mail, trackTemplates rendered, postage bought, pieces produced and tracked to the mailbox.Lob does this
Get the rep to follow upA task in the CRM timed to delivery confirmation rather than to the send button, or the whole thing lands with nobody calling.you build it
Prove it workedHold accounts back, randomize properly, match opportunities to mailed accounts, and produce a number that survives a CFO reading it.you build it

The format ceiling

Lob prints paper.
Not everything worth sending is paper.

Lob is broad within its category: postcards, letters, self-mailers, checks, all of it flat print through a press. That is a real product and it covers most of what you will send.

The ceiling is not how many formats it offers. It is that a print API cannot do a human hand or a physical object. The moment a campaign wants a handwritten note for the accounts that matter most, or something with weight in the box, you are not extending your Lob integration. You are signing a second vendor, then a third.

What each piece actually needs

Postcards, letters, self-mailersLob
Handwritten notes, in real ink2nd vendor
Gifts, with a supply chain behind them3rd vendor

Three contracts, three integrations, three invoices, and three places to enforce the same budget and approval rules. Or one.

Coverage

And not every account
is in the United States.

The same vendor arithmetic applies to geography. A print API that produces in one country either mails across the border, slowly and expensively, or you contract a printer in the next one. Here is exactly where we are, including the part that is not built yet.

United StatesPrinted and mailed domestically, at the per-piece prices published on the pricing page.live
CanadaReachable today, mailed across the border from US production, so expect longer transit than a domestic piece. Canadian domestic printing is the next provider we are adding.live
United Kingdom and the EUOn the roadmap, and honestly described as such: local production and EU data residency are the point of doing it, and neither is worth shipping half-built. Talk to us if this is a requirement rather than a preference.not yet

Every country you add is another contract, another integration, and another place the same budget and approval rules have to hold. That is the work we are taking on so it does not land on your roadmap.

The two hard ones

A model and a print API
gets you most of the way to a mistake.

The obvious version of this build is a language model writing letters into a print API. It demos well. Two things break it, and neither is fixed with more code.

01

You cannot unsend a letter

A bad email is deleted in a second and costs you nothing. A letter that gets their funding round wrong sits on a CEO’s desk for a week with your logo on it, and your rep calls into it three days later. Print is irreversible and the audience is exactly the people you cannot afford to look sloppy in front of.

That asymmetry is why every fact carries its source, why weak claims stay flagged after approval, and why nothing prints without a human signing off. It reads like process. It is the only responsible way to put generated text on paper you can never recall.

02

You have to be willing to not mail people

Incrementality requires a holdout, and a holdout means deliberately withholding the channel from accounts you believe it would help, then publishing what the comparison says even when the answer is unflattering.

Almost nobody builds this for themselves, because the team that built the program is the wrong team to run the experiment that grades it. It is not a technical moat. It is a willingness moat, and it is the reason the number at the end of a Sincerely program survives scrutiny.

The honest part

When you should just use Lob.

There is a real version of this where we are the wrong answer and the API is the right one. If you recognize yourself here, integrate Lob directly and keep the margin.

Transactional mailStatements, renewal notices, compliance letters, welcome kits. Triggered by a product event, owned by engineering, one template at volume. That is what a print API is for, and wrapping a platform around it would only add cost.
One template, no personalizationIf every piece says the same thing, the research and approval layers are overhead you do not need.
Nobody needs the numberIf mail is a brand gesture and no one is going to ask what pipeline it created, the measurement layer is the most expensive part of us and you would be paying for it unused.
You have the engineers and the appetiteSome teams genuinely should build this. If mail is core to your product rather than to your go-to-market, own it.

The cost question

Per piece is the wrong unit.

Compared per piece, we are more expensive than buying postage directly, and we will not pretend otherwise. But per piece is not the number that decides whether the channel gets funded again.

The unit that matters is what it costs to create an opportunity, and against a rep’s time the difference in postage is rounding. A channel that costs less per piece and produces nothing measurable is the expensive option.

Two ways to price the same program

cost per piececents apart
cost per opportunitythe real unit
measured against a holdoutonly one way

Published per-piece prices are on the pricing page, markup disclosed. What a pilot costs per opportunity comes out of your own holdout, not ours.

See the pricing →

Write to us

Say it on paper. Prove it in pipeline.

Launch a pilot to your top accounts: research-backed pieces your reps approve, follow-up timed to delivery, and an account-level holdout that reports what the channel actually created.

Start freeSchedule a call

Setup takes an afternoon. Delivery takes days, because paper travels. The report takes a sales cycle, because pipeline does too.

Write to us: adam@trysincerely.com

Sincerely,