A clear place for Vozubi
in a crowded market.
Vozubi should make voice and chat support easy for business owners to set up themselves, understand, and manage. Clear answers, useful customer actions, and control over costs are the core offer.
Research uses official vendor pricing, documentation, customer stories and partner pages. Prices are observations, not binding quotes; taxes, annual terms, usage definitions and extras vary. Acquisition channels are observable; their profitability and effectiveness are not public. Customer results below are vendor-published claims. Strategic recommendations and proposed experiments are our interpretation.
Selected launch pricing: Starter $39 / Pro $99 / Growth $249 are now shown on the pricing page. Review the allowances and unit economics. Checkout and live service are still in development.
Current Vozubi reality: branded marketing, local onboarding, workspace UI and browser-agent preparation tools. No live authentication, support API, recording pipeline, billing, verified installation or production integrations. Roadmap features must not be compared as shipped advantages.
Who we compete with
| Competitor / buyer | Offer and price observed | What it means for Vozubi |
|---|---|---|
| Reception by ElevenAgents Small service businesses | $29 / 75 credits, $79 / 275, $199 / 1,000 monthly. Annual totals $288 / $792 / $1,992. Web minutes use 0.5 credit; phone minutes 1; text messages 0.1. Fourteen-day trial with 30 credits. Billing documentation | The closest business-facing threat: website-scan setup, receptionist, bookings and a web widget. Compare the complete onboarding and operational workflow, not only voice quality. Quickstart |
| ElevenAgents Builders through enterprise teams | Free 15 min; regular Starter $6 / 75 min, Creator $22 / 275, Pro $99 / 1,238, Scale $299 / 3,738. LLM costs additional. Initial promotional prices differ from renewals. Pricing | Knowledge, multilingual voice and widgets are already available. Hosted OAuth MCP can manage agents, knowledge and tests from external assistants. Hosted MCP |
| Retell AI Developers, enterprises, implementation agencies | Published voice range $0.07–$0.31/min; chat $0.002+/message. Model, telephony and add-ons affect the bill. $10 trial credit. Pricing | Browser calls, tools, integrations and monitoring already exist. Its platform price omits the buyer’s implementation labor; that labor is where a focused managed product can add value. Voice API |
| Vapi Developers and technical teams | $0.05/min hosting plus speech/model components. One displayed mix estimates $82–$129 per 1,000 min. Optional Core package $29/month; Pro starts at $999/month minimum. These are not all-inclusive subscription voice prices. Pricing | Customizable voice/chat widget, tools, APIs and developer distribution. Avoid claims that putting a call button on a website is unique. Widget docs |
| Fin / Intercom Support teams, SaaS and larger organizations | Support outcomes $0.99; configured sales qualifications $9.99. Outcomes can include procedure handoffs and disqualifications. Intercom seats depend on plan/billing/promotion; Fin can accompany another helpdesk. Voice is separate, sales-assisted. Outcome definitions · Pricing | Mature support context and workflows. Fin Voice documentation describes phone deployment; that is not proof of the same embeddable web-call UX. Win on a narrower easy-to-install business workflow. Voice deployment |
| Tidio / Lyro Small businesses and ecommerce | Lyro listed monthly tiers: $39 / 50 AI conversations, $140 / 200, $350 / 500, $700 / 1,000. Helpdesk plans have separate quotas; Lyro can be purchased separately. Fifty free Lyro conversations are lifetime, not monthly. Price table · Quota definition | Low-friction written support with website integrations and handoff. Its official review says no voice support, creating a modality difference—but customers must actually value voice. Lyro capabilities |
| Zendesk Established support operations | Support Team $19/agent/month annually; Suite Team $55 annually; Suite Professional $115 annually. AI resolution billing is additional/configuration dependent. Current resolution tiers mean old flat $1.50/$2 claims are not safe universal quotes. Pricing · Resolution tiers | Channel breadth, routing, governance and a large ecosystem. Do not try to replace a full enterprise helpdesk at launch. AI agents |
| Gorgias Shopify and ecommerce teams | Helpdesk ticket-volume pricing plus AI automation. Official product facts list $1 automated interaction monthly / $0.90 annually, with $1.50 overage. Some official quota descriptions differ; verify the configured quote. Product facts · Pricing | Deep order context and commerce workflows. Voice/SMS add-ons are not by themselves evidence of autonomous browser calling. Commerce reliability and existing store context are strong advantages. Billing definitions |
Important: minutes, text messages, conversation threads, tickets and verified resolutions are different units. No percentage-cheaper claim is justified without a matched workload and feature scope. Reception overage values differ across official surfaces; obtain a checkout quote before comparing excess usage.
Customers: what buyers actually value
- Synthesia / Fin: the published story emphasizes scaling support and maintaining knowledge quality. It reports 55% resolution and handling 690% more monthly demand without adding support headcount. The lesson is operational preparation, not a guaranteed automation rate. Case study
- Cove Smart / Tidio: the story reports improved self-service and response times in product support. It presents a concrete troubleshooting problem rather than generic AI messaging. Case study
- Going / Zendesk: the startup describes entering through a startup program and consolidating tools. A subsidized beginning can seed a lasting customer relationship. Case study
- Orthofeet / Gorgias: its story reports 56% ticket automation in under two months and describes assisted implementation plus Shopify context. The takeaway is implementation assistance and measurable workflows. Case study
- Kavak / Vapi: the published story reports conversion and NPS improvements. Treat the results as specific to its deployment; they are not a voice-widget forecast. Case study
- Retell’s customer set: Anker, San Antonio Spurs, Pine Park Health and others show both enterprise and vertical adoption. This is evidence of buyer breadth, not market-share proof. Customer directory
- 3Shape / ElevenLabs: its announcement includes website FAQs and telephone support with specialist handover. Web answers plus voice escalation are already deployed elsewhere. Customer announcement
How they get customers—and what we can learn
| Observed mechanism | Evidence | Vozubi experiment |
|---|---|---|
| Low-risk entry | Tidio’s finite free AI allocation; Reception’s no-card trial; Intercom startup offer and Zendesk startup access. | Keep a capped trial. Let users test their own knowledge before paying. Track verified installation and meaningful answers, not just account creation. |
| Agencies and implementation partners | Tidio agency revenue share, Gorgias training/co-marketing, Retell solutions partners. | Recruit five WordPress agencies. Offer one reusable setup recipe, separate client workspaces and a simple monthly usage report. Calculate commission after delivery costs before offering it. |
| Optional developer and AI-assistant access | Vapi agent skills, ElevenLabs CLI, hosted agent management. | Keep direct self-service setup as the primary experience. Offer documented API/assistant access as an optional convenience for customers who already use those tools; it is not a headline promise or launch acquisition pillar. |
| Buyer-intent content | Official vendor pricing explainers, industry pages, comparison pages and documentation linked above. | Start with WordPress voice support, voice versus chat, actual cost examples and a real integration guide. Avoid mass-generated location pages. |
| Customer proof | Named case studies showing industry, workflow, rollout and claimed outcomes. | Publish three consented pilot stories with time-to-install, correct-answer checks, leads and follow-up outcomes. No invented testimonials or unsupported savings. |
| Referral and co-selling | Vapi affiliate program; ElevenLabs partner program. | Start with manually managed referrals. Pay only on collected revenue, with defined duration and exclusions. Do not build a complex partner portal before demand. |
We cannot infer paid-ad spend, CAC, organic channel share, conversion rates or partner profitability from these public pages. The experiments above are recommendations, not claims about what caused competitor growth.
Why the earlier offer needed a rethink
Not on entry price or raw voice minutes. The earlier $149 Starter offer included 200 voice minutes and 300 text-only conversations. Its $399 Growth offer included 800 minutes and 1,000 text-only conversations. Starter’s voice allocation is only 50 four-minute calls. These packages need to sell a supported business workflow, not just access to a model.
Earlier-offer workload calculator
Historical comparison using the superseded $149 / $399 offer. See the selected launch plans and their cost model for current pricing.
Starter: $149 + $0.35/min above 200 + $0.10/text-only conversation above 300. Growth: $399 + $0.30/min above 800 + $0.08/text-only conversation above 1,000. Whole conversations are rounded into voice/text counts. No tax or custom integration fees included.
| Contacts/month | Voice minutes / text chats | Starter total | Growth total |
|---|---|---|---|
| 100 | 120 / 70 | $149 | $399 |
| 300 | 360 / 210 | $205 | $399 |
| 500 | 600 / 350 | $294 | $399 |
Illustration: the 300-contact workload uses 180 Reception credits for web audio. If each of its 210 text conversations contains four billable messages, text uses another 84 credits: 264 total, within its $79/275-credit plan. This is a usage illustration, not a complete feature/service comparison. Message counts, limits and the actual customer workflow change the result. Reception unit definitions
Earlier research hypothesis: test a capped self-service offer below the $149 entry point and separate guided setup from the subscription. The selected launch pricing develops this into specific plans, usage allowances, recording retention and overages. The pricing page now reflects the $39 / $99 / $249 decision; the older figures here are retained only for comparison.
For example, a test $49 plan with 50 voice minutes and 100 text conversations has about $5.50 of assumed variable AI cost at $0.07/min and $0.02/chat. That is not gross profit: hosting, payment fees, onboarding labor, support, trial abuse and commissions remain. Use measured costs and willingness to pay to decide whether this offer works. At $149, fully used included allowances have $20 of assumed AI cost; one hour of assisted setup can matter more than that model bill.
What to build—and what not to claim
| Capability | Competitive standing | Vozubi priority / current status |
|---|---|---|
| Voice, chat, multilingual support, knowledge, recordings, summaries | Widely offered across voice platforms; not a defensible differentiator alone. | Launch essentials. UI exists; production services still needed. |
| WordPress installation and owner-friendly workspace | Website installation is common. A reliable, low-effort workflow can still matter. | Launch priority. Plugin, domain verification and real installation checks needed. |
| Verified action history with customer confirmations | Tools and analytics exist elsewhere; correctness and clarity can distinguish execution. | Build one useful workflow with authorization, idempotency and audit records before expanding. |
| Optional AI-assistant administration | ElevenLabs, Vapi and Retell already serve external AI assistants. | Browser prefill now. Add authenticated draft-first API/MCP after core self-service setup works reliably. Optional compatibility, not the main reason to buy. |
| Agency multi-client operation | Established partner ecosystems are strong. | One workspace model now; later isolated client workspaces, reusable templates, per-site limits and handover. |
| Omnichannel enterprise helpdesk, telephony, avatars | Large incumbents have breadth and mature operations. | Defer broad replacement claims, outbound dialing and elaborate avatar work. |
Our first 90 days of distribution
Start narrow: WordPress agencies serving one service-business category with valuable website inquiries. Pick the category through interviews; cleaning, landscaping or home-maintenance are hypotheses, not a claim of proven demand. Voice must outperform or complement a simple form for those buyers.
- Days 1–14: problem and activation. Interview ten owners and five agencies. Research 50 relevant businesses across two markets through public websites and directories. Score website traffic evidence, repeated questions, inquiry value and an observable support gap. Build a working install with a small approved FAQ set. No outreach is sent by this report.
- Days 15–30: three to five assisted pilots. Offer a bounded pilot with a defined success criterion and usage cap. Observe setup and handoff. Track knowledge-backed answers, captured qualified inquiries, failed calls, human escalations and variable cost.
- Days 31–60: proof and reusable setup. Publish an installation guide and consented case studies. Create one reusable industry configuration. Test self-service versus guided pricing with actual purchase intent. Add partner referrals only after understanding support workload.
- Days 61–90: retention, referrals and repeatable installs. Review retained pilot customers and the support work they actually saved. Improve the self-service steps where owners struggle. Ask successful customers for referrals and help partners repeat verified installations. Expand only channels producing retained paying users; keep optional AI-assistant documentation available for customers who prefer that workflow.
Weekly scorecard: qualified prospects demos workspace setup verified installation first useful conversation paid activation retained customer. Record acquisition source, founder time, infrastructure/AI cost, refunds and support time. Treat conversion targets as experiments; do not mistake account signups for demand.
Discovery versus actual product use
AI discovery needs accessible, accurate product pages, pricing, citations and independent evidence. Google explicitly says special AI schema is unnecessary and llms.txt does not improve its search ranking. Our llms.txt serves tool/document navigation, not a recommendation guarantee. Google guidance
People should be able to set up and manage Vozubi directly through the dashboard. For customers who choose to delegate tasks to an AI assistant, optional product access requires authenticated, scoped capabilities with reliable results. Build one business API on Railway for the dashboard, then expose the same services through optional MCP access. Never let an assistant turn a draft into live spend or a financial action solely because it filled a form. See the agent guide for what works now; the source plan documents the production design.
Visibility limit: the current Sites deployment remains owner-private. Crawl files do not bypass that access control. For launch on Vercel, choose the public domain, update canonical/sitemap URLs, verify search crawlers can reach marketing content, and keep the app authenticated. Move this founder report and strategy material behind authentication before making the site public.