Product guide
Recommendations Engine
What it is
Most of GlassDollar starts with you: you have a pain point, you search, you source. The Recommendations Engine works the other way around — it’s a newsletter, not a search tool. Instead of waiting for a sourcing request, it proactively lands in your inbox with a small number of curated startup and technology recommendations, based on what’s already worked for comparable organizations.
Anyone on your team who wants it can receive it — it’s not limited to System Admins. If you want in, or want to add a colleague, ask your account manager.
Each send is short on purpose — a couple of picks, not a list. Every recommendation is anchored to a specific problem or space, backed by a credibility signal (how many startups map to it, how many peers have already procured in that space), and links straight into SourceAI — so a recommendation that catches your eye turns into a live search in one click, not a bookmark you forget about.
| Type | What it means |
|---|---|
| Frontier space | A space your organization isn’t in yet, but comparable peers are starting to move into — worth watching before it’s obvious |
| Proven pick | A solution type with a track record — other industrial peers have already run PoCs or procured in this space |
How recommendations get chosen
Recommendations are matched using two inputs — one about the world, one about you:
- Public signals and anonymized deal data. What your organization (and comparable peers) have publicly announced, scouted, or procured, cross-referenced against six years and 200+ Venture Clienting programmes’ worth of sourcing and PoC outcomes.
- Your declared focus areas. The specific strategic and innovation fields you tell us matter to your team — see below. This is the single biggest lever you have over how relevant your recommendations are.
On data: recommendations are matched on public organizational signals and anonymized, aggregated patterns from across our customer base — never your internal project details, pipeline data, or anything confidential to your organization.
Tell it what matters to you
The engine is only as good as what it knows to look for. If your focus areas aren’t set, recommendations default to broad, public signals about your organization — useful, but generic. Tell it your actual strategic fields, and the picks get a lot sharper.
Submit your focus areas: gd-recs-feedback.netlify.app/focus
| Step | What happens |
|---|---|
| Open the focus link | A short form asking which strategic and innovation fields matter to you right now |
| Fill it out — for yourself or your team | Submit as an individual, or fill it out on behalf of colleagues whose focus areas you already know |
| Recommendations recalibrate | Future sends are matched against your declared fields, not just public signals |
Ask your account manager for a personalized version of this link, pre-filled with your organization and email — one less field to fill in.
Cadence & format
Cadence depends on who’s on the list:
| Recipient | Typical cadence |
|---|---|
| Admins | Weekly |
| Broader teams and business units | Monthly — easier to keep up with, and gives each pick room to actually get looked at before the next one lands |
| Anyone by specific agreement | Whatever’s agreed with your account manager — cadence isn’t fixed once you’re on the list |
If the cadence you’re on isn’t working for you or your team, your account manager can adjust it in either direction.
A German-language version is available on request. As more teams set SSO and individual focus areas, recommendations are moving toward department- and role-level personalization — so different stakeholders on the same account can get picks relevant to their own remit, not one generic send for the whole company.
Common questions
| Question | Answer |
|---|---|
| Who can receive the newsletter? | Anyone on your team who wants it — not just System Admins. Ask your account manager to add yourself or a colleague. |
| Does this use our internal data? | No — only public signals about your organization and anonymized, aggregated patterns from across our customer base. |
| Can I stop receiving these? | Reply to any recommendation email, or contact your account manager to opt out. |
| Can different colleagues get different picks? | Yes — submit focus areas per person or department via the link above, and ask your account manager about setting up tailored sends per stakeholder. |
| A recommendation didn’t fit — does that feedback go anywhere? | Yes — every recommendation has quick feedback buttons, and what you flag as “not for us” helps steer future sends away from it. |
Related pages
- SourceAI Search
- Decentralization via Business Units — using the Recommendations Engine’s focus fields and a monthly newsletter to drive early interest in a newly decentralized business unit