Customer onboarding email series with an avatar: 5 clips, cost by tier

Five 20-second onboarding clips from one Sume avatar cost $18.40 on standard, $24.50 on plus and $55.00 on max. How to plan the series and when to re-render.

4 min readSume
All posts

A five-email onboarding series with one 20-second avatar clip in each email is 100 seconds of avatar video, and it costs $18.40 at standard, $24.50 at plus, and $55.00 at max, with no product image. Add the one-time $0.95 to create the presenter if you do not already have one. Sume avatars are async jobs: you submit a request, get a job id back, and read the finished video later. There is no live video session.

Onboarding is a good fit for pre-rendered clips because the sequence is known in advance, the same clip goes to every new user, and the content changes only when the product does.

The series and its price

Each clip is one job. Each job is between 4 and 60 seconds, and 20 seconds is a comfortable length for one idea per email. The rates are the no-product rates from provider-pricing.

Five 20-second onboarding clips by tier (provider-pricing, read 2026-10-07)
EmailTopicStandardPlusMax
1Welcome and first step$3.68$4.90$11.00
2Connect your data$3.68$4.90$11.00
3Invite a teammate$3.68$4.90$11.00
4Your first report$3.68$4.90$11.00
5Where to get help$3.68$4.90$11.00
Total100 seconds$18.40$24.50$55.00

One avatar, five jobs

Create the presenter once. Avatar creation is a flat job that returns a reusable handle, and every talking video uses avatar_handle to point at it. Using the same handle in all five clips keeps the face and voice consistent, which is what makes a series feel like one product and not five unrelated clips.

Submit the five jobs together if you want them ready at the same time. They are independent jobs, so they can queue and run in parallel up to your workspace limits. A paid generation can sit in queued before a worker takes it, which the docs describe as a normal accepted state.

for n in 1 2 3 4 5; do
  curl -s -X POST https://api.sume.com/v1/avatar-1.0/talking-video \
    -H "Authorization: Bearer $SUME_API_KEY" \
    -H "Content-Type: application/json" \
    -H "Idempotency-Key: onboarding-2026-10-email-$n" \
    -d "$(cat email-$n.json)"
done

Keeping the series current

When the product screen changes, only the clip that talks about it needs a new render. Keep each script in version control with the job id and tier next to it, and re-render one clip at a time. Re-rendering email 3 at plus is $4.90, not the $24.50 for the whole series.

Put the clip behind a thumbnail in the email and link to a page that plays it. Use the media URL from the job result, which is a public media.sume.com URL, rather than a raw provider link, because the docs say provider URLs are not public API outputs.

Budget guardrails

The submit reserves the estimated cost from the workspace balance, so a request with too little balance returns 402 insufficient_credits before any provider work starts. For a series, that means you find out at submit time, not after the third clip. A quick approach: render email 1 at your final tier first, approve it, then submit the other four.

Writing for email

Email changes how people watch. Most will see a thumbnail and decide in a second whether to press play, so open each script with the action the viewer will take, not with a greeting. "Connect your first data source in under a minute" beats "Hi, welcome to the product". Keep sentences short, because the avatar reads at a steady pace and a 20-second clip holds only about 50 words.

Add captions. Many people open email with the sound off, and burned-in captions make the clip usable anyway. The captions option adds styles to the clean final MP4 after generation, using the spoken script. The caption stage is a soft failure by design: if it fails, the avatar job can still succeed with a clean video, and you can caption again with the standalone captions route.

Tier choice for a series

Since the same five clips go to every new user for months, quality pays off more here than in a one-off. A reasonable plan is standard while the scripts settle, then a final pass at plus, which is the default tier when you leave quality out. The difference for the full series is $6.10, which is small next to the cost of a series that looks rough in front of every new customer.

Sources

Related posts

More in Use cases

All Use cases posts

Written by Sume