Almost every deal under 300 workers, and most above. A group per site, run from the site manager's personal handset, often with a second group for the subcontractors. Usually raised early and cheerfully, because they think it settles the conversation.

All of that, mostly. It is free, it is instant, adoption is universal and the crews genuinely like it. Nothing you sell will be faster to send a message with, and you should say so. WhatsApp also shipped on-device message translation in September 2025, so "it doesn't translate" is no longer accurate on its own.

• Translation is sender-side. You publish once; every worker reads it in their language, captions included. WhatsApp's is a reader action, long-press and Translate, per chat, and it says "translations occur on your device where WhatsApp cannot see them", so nothing is reported back to a sender. Its help page adds the feature "is currently not available on the WhatsApp Business App".
• Ticks are not a record. Recipients can switch them off and they cannot be exported. A tick means displayed on a device, not read by a person, and says nothing about the people never added to the group.
• Broadcasts fail silently. WhatsApp's own words: "Only contacts who've added your number to their phone's address book will receive your broadcast message." A list holds 256, and a monthly broadcast limit is rolling out in some countries.
• Offboarding. When the site manager leaves, the group, the history and every worker's mobile number leave with him. A departed worker stays in the group until someone remembers. TeamTalk revokes on a leaving date and keeps the record.
• Data protection. The group makes a personal handset a processing location for personal data about every worker in it. You remain the controller and carry the liability, with no access, no retention control and no way to answer a subject access or erasure request.
• Doing it properly is not free. The WhatsApp Business Platform has been priced per message since 1 July 2025 and needs a business solution provider, template pre-approval and a developer. A daily toolbox talk is unlikely to qualify as a "utility" template.

Any organisation with a Microsoft estate and an IT function in the room. It rarely comes from the HSEQ lead, who knows perfectly well that the bricklayers are not in Teams. It comes at the point where the deal has to be approved.

Teams is a good product and they do already pay for it for office staff. Microsoft 365 F1 is a genuine frontline SKU at £2.30 per user per month, and F-SKUs have no seat cap or minimum. The "500-seat rule" that circulates relates to Enterprise Agreement minimums, not F-SKU eligibility: do not repeat it, because an IT manager will correct you. Read receipts and message translation both exist.

• Read receipts are structurally unfit as a compliance record. Microsoft's own documentation: they work in "1:1 and group chats 20 people or fewer", they do not exist in channels at all, the default policy is user-controlled so a worker can unilaterally switch off your proof, and they "aren't captured in eDiscovery reporting".
• Translation is recipient-side and opt-in. The "Translate messages" policy translates into the language set in the user's own Microsoft 365 profile. Admin-gated, user-configured, no sender-side guarantee and no per-recipient confirmation.
• Every worker needs a provisioned identity and a paid licence. There is no unlicensed participant model. For agency labour, subcontractors and three-month site staff that is an IT project per person, plus password and MFA support for a population that has never had a corporate login. TeamTalk is a QR code and twenty seconds.
• The real number. The cheapest credible Microsoft stack for tracked frontline comms is F1 at £2.30 plus Viva Employee Communications and Communities at £1.54: £3.84 per user per month on an annual commitment. For 500 workers that is roughly £23,000 a year committed, before the identity project, and it still does not deliver proof of receipt.
• Viva Engage is never standalone. It is a communities product and always an add-on on a base Microsoft 365 seat, so the licence question does not go away.

Three hundred to five thousand multilingual workers, usually where somebody owns internal communications. UK-founded, with a strong UK logo wall including Stagecoach, Go-Ahead, First Bus, National Express, JD Sports and Costa Coffee, and it holds Cyber Essentials Basic and Plus, ISO 27001 and SOC 2.

Mandatory Reads turn a document into a sign-off that staff must actively acknowledge; acknowledgement data pulls into Excel or CSV; reminders can be targeted only at the people yet to confirm. Activation works by phone number or a shared group code with no email. Near-miss forms are unlimited at no extra cost. Offline mode is real. The critical comms page claims 40+ languages. It is the benchmark in this category and you should say so before you differentiate.

• Reporting and export may be a paid add-on on Blink; confirm against their current plan. Blink's published pricing lists "Reporting and exporting data" as a separately priced add-on on all three tiers, so the compliance artefact you are actually buying may cost extra. Say it as a question to put to them, never as a fact. Ours is included on every plan.
• No video captions at all. Not translated, not even auto-generated. A site manager's forty-five second phone video about a changed lifting plan is the format sites actually use, and captioning it in each worker's language is the one thing nobody else in this market does.
• No QR onboarding. A shared group code is the nearest equivalent. There is no printable join card that gets a whole site enrolled in one break.
• Multilingual sites pay the top rate. Advanced multi-language publishing sits in Pro at £3.75, so the site with five languages, the one that needs this most, pays the highest published price.
• No UK data residency. Blink's help centre names its regions: AWS Ireland, Sydney and Oregon; no UK region (Blink help centre, checked 28 Aug 2026). Its privacy policy states data is transferred outside the UK and EEA, including to the US. Ours is the United Kingdom (AWS London, eu-west-2). For NHS trusts, councils, care providers and unionised accounts this is a live procurement question.
• Rollout. Blink's own competitor guide gives two to six weeks for a typical SMB rollout, and six to twelve weeks for larger or multilingual ones.

Every construction and facilities conversation, and most catering ones. The sheets live in a lever-arch file in the site cabin, or as photographs in somebody's camera roll, or scanned into a folder on a shared drive that nobody has opened since the last audit.

It is genuinely evidence, it is accepted practice, it costs nothing to start and it works when the power is off and there is no signal in the basement. Take it seriously: the HSEQ lead in front of you almost certainly designed the form, trained the supervisors on it and defended it in an audit.

• The sheet is in English, and so was the talk. A signature proves attendance, not comprehension, and regulation 10 of MHSWR 1999 asks for information that is comprehensible. The question that lands: "what language was the sheet in?"
• Finding one, a year later. Produce every briefing a named operative received on a named date, across four sites, in under a minute. A lever-arch file cannot. A CSV export can, and so can a printable per-worker record.
• The chase. A sheet records who turned up. It never records who did not. The reach page hides everyone who has confirmed and leaves you the outstanding list, and reminders go only to that list.
• The people who were never in the room. Late starts, night shift, agency cover, the subcontractor who arrived at eleven. They get the notice on their own phone, in their own language, and it is recorded.
• Everything between the talks. A signature captures one briefing. Every rule change, permit suspension and revised muster point in between is undocumented.
• Wet, lost and left behind. Sheets go missing, ink runs, folders leave with the manager. A record held once, centrally, does not.

Two thousand employees and up, with an internal communications function, or any conversation that has drifted above a division to group level. Sometimes it arrives as a rumour: "group are looking at something." Find out whether it is signed or still a slide.

Staffbase is a mature enterprise intranet plus employee app, employee email, SMS and digital signage: 1,500+ enterprise customers, 70+ languages, built-in translation, incident reporting and no-email login. Read confirmation is not documented publicly, which is not the same as absent, so do not claim they lack it.

Workvivo (now part of Zoom) has a strong brand, Auto-Translate, the best subtitle controls in the category, and a documented Read & Acknowledge feature with a who-has-and-has-not report, reminders and export. Say so before you differentiate.

• Workvivo is priced so the frontline is the part that never gets licensed. Workvivo documents a Read & Acknowledge feature: admins flag content as a mandatory read, users acknowledge, the admin gets a report of who has and has not, with reminders and export. Do not claim it is missing. The argument is the price: its published floor is $20,000 a year for 250 to 2,000 employees, so a 400-person contractor licenses head office and the 300 operatives the notice is for stay outside it. Its subtitles are auto-generated but not auto-translated per viewer.
• Entry price. Workvivo's own site publishes the floor: "$20,000 yearly for companies with 250-2,000 employees". Staffbase publishes no price list at all; third-party benchmark data puts the median annual contract near $28,500, with implementation from $15,000 to $75,000 on top.
• Time to go live. Staffbase's own pricing page says 90 days. A site that mobilises next month cannot wait for that, and neither can the notice that has to go out on Monday.
• UK data residency. Neither offers it. Staffbase is explicit: EU (Azure Frankfurt), US (Azure Virginia) or Australia only. Workvivo publishes no data residency at all.
• The buyer. Both are sold to an internal communications department. A three-hundred-person contractor does not have one, and the HSEQ lead who actually has the problem has no route to that budget line.

Principal contractors and housebuilders that have already bought a site safety platform, usually through the HSEQ function rather than IT. Present in the UK, Australia, New Zealand and the US. If they mention orientations, JHAs, permits or subcontractor management in discovery, assume it or something like it is there.

HammerTech's Safety Meetings module runs toolbox talks, pre-starts and safety briefings from one platform with attendance tracking and sign-offs. It covers subcontractor management, orientations and inductions, JHAs, safety data sheets, pre-task plans and permits. It prices on annual construction volume plus the number of active job sites, explicitly not per user seat, and subcontractors get free access. It is a serious product with a real construction model, and pretending otherwise will end the meeting.

• It does not translate. HammerTech proves attendance at a toolbox talk. It does not put the talk in front of the operative in a language he reads. The question to ask, once, quietly: "what language was the sign-off sheet in?"
• Attendance is not comprehension, and the regulation asks for comprehensible. Regulation 10 of MHSWR 1999 requires "comprehensible and relevant information". A signature on an English form from a worker who reads Romanian is evidence of attendance and nothing more. Section 40 puts the burden on the employer to prove it did what was reasonably practicable. A per-person, timestamped, language-stamped confirmation is evidence for that. It is not a defence on its own.
• It is a meeting record, not a channel. The permit suspension at 11am, the revised muster point, the lifting plan change on Thursday: those are not toolbox talks and they do not have a meeting to attach to. That is the traffic TeamTalk carries.
• No translated video captions. A site manager's forty-five second phone video, captioned in each worker's language, does not exist anywhere else in this market.
• Price. HammerTech publishes no figures at all. We publish ours: £180 a site a month, £2,160 a site for the year, everyone on it. A published price against an unpublished one is an advantage; use it without disparaging them.

Almost any organisation of 300 or more with a QHSE function: construction, facilities management, catering, logistics. Usually already bought, already loved, and already the budget holder's own decision. Strong UK presence with a Manchester office. Priced in US dollars, which finance will notice.

Inspections and audits, the old iAuditor, are the best in the category. Heads Up shares a video, image or PDF instead of a wall of text, targets a single site or group, shows who received it and who needs a follow-up, and captures digital signatures on critical documents. Issues lets contractors and visitors report from their own phone by QR code with no login. Free up to 10 seats. Premium is $24 a seat a month billed annually, with site-based pricing available at Enterprise.

• At $24 a seat, the operatives are never licensed. That is the whole opening and it is arithmetic, not opinion. A 400-person contractor buys 35 seats for supervisors and QHSE and leaves 365 people outside the system. Heads Up then structurally cannot reach the people the toolbox talk is for. Ask: "how many seats do you have, and how many people are on your sites?" Let them do the subtraction.
• Translation sits in the wrong module. Mitti's AI narration and translation into 100+ languages is documented in Training. Nothing is documented for translating a Heads Up broadcast per recipient. Ask them to demonstrate a Heads Up message arriving in Polish; do not assert that it cannot.
• No phone-only login for unlicensed workers. Issues can be raised by QR with no login, which is genuinely good, but receiving and confirming a mandatory notice is a licensed action.
• No translated video captions. Same gap as everyone else in this list.
• Price and currency. £180 a site a month, £2,160 a site for the year, in sterling, everyone on the site included. Mitti Premium is roughly £19 a seat a month at current rates, and it is a dollar invoice.

Every deal, at the end. It rarely arrives as an objection. It arrives as agreement: they like it, they can see it working, and then nothing happens for four months. The tell is enthusiasm with no date, no owner and no next meeting in the diary.

It costs nothing to keep, it has never yet failed in a way anyone noticed, and it is what got them through the last audit. The supervisors know how to work it. Changing it costs somebody real effort and earns them nothing visible. Every one of those is a good reason to do nothing, and none of them is a reason it is safe.

• Make the cost of doing nothing specific, once. Not a scare story: one question. "If you had an incident on Tuesday and the inspector asked what that operative was told and what language it was in, what would you hand him?" Then be quiet. If the answer is a lever-arch file and a WhatsApp group, they have heard themselves say it, which is worth more than anything you can say.
• Shrink the decision until it is not a decision. Sixty days, three sites, a fixed £2,500, refunded if both of the first two criteria are missed, and fully credited if they go ahead. There is no way to lose money, so "let's wait" stops being the cautious option.
• Give it a deadline that is theirs, not yours. A mobilisation date, an audit, a PQQ, a principal contractor's onboarding, a client requirement. Deals move against the customer's calendar and never against yours.
• Name the person who has to do the work, and make it small. One administrator per site, two posts a week, a printed card on the cabin door. If nobody on their side owns it, it will not happen however good the product is.
• Always leave with a date. Never end on "I'll send you some information". Thirty minutes with the site manager, on a real phone, Tuesday or Thursday.
• Give them something to forward. The one-page internal note for their colleagues does the arguing when you are not in the room, which is most of the time in a four to seven stakeholder decision.