Marcus Norrving
CEO & co-founder at Garba · Updated August 2026
Garba and Attention are solving the same problem — get the admin out of the rep's day by turning every call into CRM updates, follow-ups and coaching. Both do it well, and neither is a notetaker.
Choose Garba if you're European and want EU data processing to be the default rather than something you have to establish in procurement, or if rep adoption is the thing that failed last time.
Choose Attention if AI-graded scorecards on every call are the point, if you want a wide integration surface and LLM-native extensibility to build on, or if you're wiring sales data into custom workflows.
| Attribute | Garba | Attention |
|---|---|---|
| Best for | European sales teams that need adoption, not another dashboard | Teams that want every call graded, and a platform to build on |
| Built around | The rep's day — output arrives without the rep going to get it | The coaching loop — graded calls feeding a compounding context layer |
| EU data residency | Default. 100% of data processed in the EU, nothing to configure | Not published. Their security page lists SOC 2 Type II and HIPAA, and names no data location — confirm before signing |
| CRM auto-fill | Field-level from every conversation — no rep data entry | Automatic after every call |
| Follow-up drafting | Recap and next steps drafted the moment the call ends | Drafted after calls |
| Coaching scorecards | Reps self-score against your playbook, straight after the call | AI-graded on every call, with configurable frameworks — deeper here |
| Integrations | Native HubSpot, Salesforce, Pipedrive, plus Zoom/Meet/Teams/Slack, MCP server, REST API and webhooks | 200+ integrations, plus a Claude Connector and an MCP server — wider surface |
| Language coverage | 125 languages, with accent-sensitive detection | 100+ languages |
| Pricing | Custom | Reported ~$59–200/user/mo — not published |
GDPR and data residency
Garba
Garba processes 100% of customer data in the EU. It's structural, not a region you select at onboarding or an enterprise add-on, and a DPA is provided as standard. Garba AI AB is a Swedish company operating from Malmö.
Attention
Attention's security page states that it is SOC 2 Type II compliant and HIPAA compliant, and describes encryption in transit and at rest. It does not name where customer data is stored or processed, and does not mention GDPR.
That is a statement about what's published, not an accusation — they may well have an answer that satisfies your security review. But if you're recording EU customer conversations, "not stated on the website" is where the review starts rather than ends. Ask for the data location and the DPA in writing before you sign.
Who the product is built around
Garba
Garba's premise is that the rep gets paid first. The CRM is updated without typing, the follow-up is drafted before the rep leaves the call, and the playbook score arrives without a manager in the loop. The rep-facing surface is deliberately small, because value that requires a behaviour change is value most teams never collect.
BIMobject replaced 50+ Gong licences with Garba in 2025 and had adoption across the revenue team in week one — after Gong had never landed.
Attention
Attention's centre of gravity is the coaching loop: calls get graded, reps engage with the feedback, and the system's context improves as more calls and outcomes feed in.
That's a genuinely strong model, and where coaching is already a management habit it compounds. It does ask more of the rep than "the work is already done", so how well it lands depends on the coaching culture around it.
Coaching — Attention's strongest card
Garba
Garba scores each rep against the team's playbook immediately after the call, so the feedback lands while the conversation is still fresh and doesn't wait on manager capacity. Managers get the aggregate picture without reviewing calls one at a time.
Attention
Attention grades 100% of calls against scorecards you write in plain language — standard frameworks like MEDDIC and BANT, or your own. Every call is scored, not just the ones a manager picked out.
Straightforwardly: the scorecard infrastructure is more configurable and higher-volume than ours. If coaching is the primary lever you're pulling on win rate, that's the stronger tool, and we'd rather you knew it here.
Integrations and extensibility
Garba
Garba integrates natively with HubSpot, Salesforce and Pipedrive, plus Zoom, Google Meet, Microsoft Teams and Slack, and ships an MCP server, a REST API and webhooks for everything else. Deep on the systems a sales team actually lives in, rather than broad across a long tail.
Garba isn't built to be a platform layer. It does three jobs — CRM hygiene, follow-ups, coaching — and is opinionated about them.
Attention
Attention connects to 200+ tools and offers a Claude Connector alongside its MCP server, so it composes into LLM-native workflows: querying pipeline from Slack, feeding custom agents, wiring sales data into whatever you're building.
If your GTM stack is wide, or you're building custom AI workflows on top of your sales data, that surface is a real advantage and the honest reason to pick them.
Language coverage
Garba
Garba transcribes 125 languages with accent-sensitive automatic detection. Worth stating plainly because this comparison is often written the other way round: Garba is built for European sales teams, which is a statement about the sales motion, the data-residency expectations and the length of the cycle — not a limit on which languages it understands.
Attention
Attention supports 100+ languages. For a globally distributed team calling across many markets, both tools clear the bar; this is not the row that decides it.
Price
Garba
Garba's pricing is custom. The concrete reference point we can give: BIMobject came in at roughly a third of their previous Gong contract, for a tool the whole team then used.
Attention
Attention does not publish pricing. Reported figures land around $59–200 per user per month depending on tier and seat count — treat that as reported rather than quoted, and get your own number.
Choose Garba when — you're in Europe and want EU processing as the default rather than a procurement conversation; rep adoption is what failed last time; you want CRM hygiene, follow-ups and coaching handled without a behaviour change; you want it live in days.
Choose Attention when — graded scorecards on every call are the primary lever; you need a wide integration surface or Claude/MCP extensibility; you're building custom AI workflows on your sales data; your team spans many markets and a platform layer is the point.
Attention's security page lists SOC 2 Type II and HIPAA compliance and describes encryption in transit and at rest, but does not state where customer data is stored or processed, and does not mention GDPR. That means it isn't published — not that it doesn't exist. If EU data residency is a requirement for you, ask Attention directly for the data location and get it into the DPA before signing.
Yes. Garba processes 100% of customer data within the EU by default, with nothing to configure, and a Data Processing Agreement is provided as standard. Garba AI AB is a Swedish company based in Malmö.
Attention, if coaching depth is the deciding factor. It grades every call against scorecards you can write in plain language and configure per framework, which is more configurable and higher-volume than Garba's approach. Garba has reps self-score against the team playbook straight after the call, which gets feedback in front of the rep faster and without manager capacity, but is less configurable. Pick on which of those two properties you actually need.
125, with accent-sensitive automatic detection. Comparisons sometimes describe Garba as English- and Swedish-only because it is built for European sales teams — that's a statement about the sales motion and data-residency expectations it's designed around, not about language coverage.
Neither publishes pricing. Attention is reported at roughly $59–200 per user per month depending on tier and seat count. Garba's pricing is custom; the concrete reference we can point to is BIMobject, who replaced their Gong contract with Garba at roughly one-third of what they had been paying. Both numbers are reported rather than quoted — get your own.
Yes, both write CRM fields from call data with no manual rep input. The difference is downstream: Garba routes the output into the rep's workflow (drafted follow-ups, playbook scores), while Attention routes it into scorecards and a context layer that compounds across calls.
Marcus Norrving
CEO & co-founder at Garba
Co-founded Garba after helping build Notified, the PR and investor-communications platform acquired by West Corp in 2019 - hiring the reps, running the pipeline, and writing the 9pm follow-ups. Started Garba in Malmö the moment AI could do that work instead of just recording it.
We're a competitor writing about a competitor. Every Attention claim here comes from their own public pages, and where they're stronger we've said so. Something out of date or wrong? Email support@garba.ai and we'll correct it.
Twenty minutes, your CRM, your last real call. Book a demo.