How to choose an IoT SIM

Picking the wrong SIM at project start costs money and usually delays launch by months. This guide walks the decisions in the right order, starting from your use case and ending at the tariff. Each step keeps the essentials and leaves detail to the glossary and comparison pages.

1. 1

   Define device mobility and latency

   If the device moves or needs sub-second server response, rule out NB-IoT. For static indoor, NB-IoT is a candidate. For everything else, LTE-M or Cat-1 bis.

   Tip: Ask product for a real location map: coverage is what breaks pilots, not specs.

2. 2

   Pick the radio tech

   Cat-1 bis for general international, LTE-M for LPWA with mobility, NB-IoT for static LPWA, standard 4G if you need bandwidth. 2G only when forced by legacy hardware, and accept the sunset.

3. 3

   Decide the physical form

   MFF2 solderable for industrial, automotive, outdoor. 2FF/3FF/4FF removable for pilots or where the customer demands access. Above 5k units, consider eUICC from day one.

4. 4

   Define network needs

   Static IP needed? Does the server initiate the connection? If so, private static IP with private APN. Outbound only, public APN suffices.

   Tip: Every inbound connection is a security decision. Document who can initiate and why.

5. 5

   Map geographic coverage

   List deployment countries. If any restrict permanent roaming (BR, TR, IN, CN, RU), plan eUICC or Multi-IMSI from the start.

6. 6

   Size the data plan

   Estimate MB/month per device (with 30% margin) and multiply by fleet size. For high volume, pooled plan. For very variable use, pay-as-you-go.

   Tip: Ask for sample CDRs during the pilot: do not trust your estimates, measure.

7. 7

   Vet the portal and API

   The provider must offer activation/deactivation, usage alerts, CDRs over API and, to scale, multi-user and RBAC. Without this, the carrier is just a reseller.