Top 10 RFID Implementation Mistakes (and How to Avoid Them)
Every RFID programme we get called into after the fact has made one of the same ten mistakes. Some of them are subtle, some are obvious only in hindsight, all of them are avoidable. This post is our internal checklist, shared publicly — read it before you kick off a pilot, not during the firefight afterwards.
Mistake #1 — Starting too big
The single most common failure mode: a CIO buys the pitch, approves a pan-India rollout, and the first "pilot" is three warehouses across three cities simultaneously. RFID has too many moving parts — tags, readers, antennas, antenna tuning, Wi-Fi, software, integration, staff training — to debug all three locations in parallel.
Fix: One site. One zone within that site. Get it running cleanly for 4–6 weeks before you copy the blueprint to Site 2. Every deployment we've succeeded with started with a single-zone pilot; every one we've rescued started with "let's do everything at once."
Mistake #2 — Wrong tag for the material
A standard UHF inlay on a metal pallet reads at 0%. A 13.56 MHz HF tag on a warehouse carton reads with 30% reliability because the range is too short. These aren't edge cases — they're the most common tag-selection errors we see.
Fix: Match the tag to the material and use case:
- Metal assets → on-metal UHF tag
- Plastic / wood / cartons → standard UHF inlay label
- Library books → HF library tag
- Linen / uniforms → washable laundry tag
- Access cards / patient wristbands → HF / NFC
The decision framework in our UHF vs HF vs NFC guide walks through this in more depth.
Mistake #3 — Under-tagging
Tagging 70% of cartons gets you 0% of the value. Because the 30% of untagged inventory breaks every count, and staff stop trusting the system.
Fix: Commit to 100% tagging before you cut over. If 100% isn't achievable in one go, tag by category completely — "all finished goods" or "all IT assets worth ₹25,000+" — rather than randomly tagging across categories.
Mistake #4 — Parallel barcode scanning
"RFID is new, let's also keep scanning barcodes just to be sure." Now you have 2× the labour and 1× the benefit. This kills ROI on paper and kills morale on the floor.
Fix: Cut over cleanly after a 2-week overlap. Management has to actively stop the barcode process. Otherwise the barcode process will quietly win.
Mistake #5 — Ignoring the read environment
UHF reads are sensitive to metal shelving, concrete walls, water tanks, forklift charging bays, Wi-Fi APs, and neon signs. We've seen pilots fail because the "clean" floor map had a stack of steel drums 2m from the portal reader.
Fix: Do a site survey with a test reader before you order hardware. Walk the aisles with a handheld, observe read rates, identify dead zones. Adjust antenna placement or pick different reader models.
Mistake #6 — Reader placement as an afterthought
Gate readers get installed wherever the electrician found a power outlet. The result is off-axis reads, false triggers, and missed tags.
Fix: Plan reader placement with the RFID integrator, not the electrician. Readers belong where goods must pass, at the right height, with clear line of sight. Our reader buying guide has placement recipes by use case.
Mistake #7 — No software owner after go-live
Day 1 everyone's excited. Day 60, nobody knows who owns the RFID platform. Master data drifts. New assets go un-tagged. Staff changes their process without updating the system. In 12 months you've got a ₹50 lakh system being used as a glorified barcode replacement.
Fix: Assign a single RFID Programme Owner on the customer side before signing the purchase order. Not IT. Not Operations. A dedicated person with floor-and-systems credibility.
Mistake #8 — Skipping integration
RFID sits on a tablet in the corner showing pretty dashboards, while ERP, WMS, and the ecommerce site run off separate stock counts. The data never reaches the decisions it should influence.
Fix: Scope integration from day one. Identium's platform ships REST APIs and webhooks; a typical ERP integration is a 2–3 week project. Skip this and you've built an expensive isolated island.
Mistake #9 — Ordering hardware without testing one unit
₹15 lakh of readers arrives. Only then does someone discover the firmware doesn't support the chosen tag chip family, or the handheld's Android build doesn't install your corporate MDM.
Fix: Buy one unit of each reader/tag model first. Validate with the exact software stack, over the exact Wi-Fi network, with the exact tag you'll use in production. Only then place the bulk order.
Mistake #10 — Under-investing in training
Staff who don't trust the system, don't use the system. Staff who don't use the system generate no data. Staff who generate no data are invisible to the dashboards that justified the project.
Fix: Budget for 4 hours of hands-on training per operational staff member, plus follow-up floor walks in weeks 2 and 3. Get the confident users to coach the hesitant ones. Make the trainers visible and accountable.
The 11th mistake (bonus): treating RFID as IT
RFID fails when IT runs it. It succeeds when operations runs it, with IT support. The daily stakeholders are warehouse associates, shop-floor supervisors, nurses, or event staff — not systems admins. Build for them.
If you remember one thing from this post
Start small, tag completely, pick the right hardware for the material, and put a single accountable owner on it. The other six items in this list basically follow from those four.
If you're planning a deployment and want a second pair of eyes on your tag selection or reader mix, we do free 30-minute consultations. The common-sense tag/reader checks catch most of these mistakes before the purchase order goes out.
