Questions on AI go-to-market, answered
Short answers from Aarya Ravahi on go-to-market, AI in sales, and selling into the Gulf, with links to the full pieces.
Who is Aarya Ravahi?
Aarya Ravahi is an AI go-to-market leader and independent consultant based between Dubai and London. Ten years in revenue across enterprise software, AI and Web3, including Celonis, SUSE, Gelato, Parity (Polkadot) and Galadriel, where an AI infrastructure product went from launch to $180k in monthly recurring revenue in four months. More about Aarya
What experience does Aarya Ravahi have in AI sales and go-to-market leadership?
Head of GTM at Galadriel, an AI infrastructure company. Head of Growth at Gelato Network. Senior Growth Manager at Parity (Polkadot). Quota-carrying enterprise sales at Celonis and SUSE. Now an independent AI go-to-market consultant. Suited to Head of Growth, Head of GTM, VP Sales and GTM engineering leadership roles in the UAE, the wider Gulf and the UK. Full experience
Where does Aarya Ravahi work?
Between Dubai and London, with work across the UAE, Saudi Arabia and the wider Gulf, the UK and Europe. Earlier roles were based in Amsterdam, London, San Francisco and Dubai. Most work is remote, with time on site where it helps. Get in touch
How can a B2B sales team use AI to grow pipeline?
Use AI to find and research accounts from buying signals rather than broad lists, draft first messages that are specific to each account, and take call notes, CRM updates and reporting off reps entirely. The gain is more well-researched conversations per rep, not more emails. AI enablement for revenue teams
What are the limits of using AI to grow pipeline?
AI makes activity cheap, so it is easy to fill a pipeline with conversations that will never close. It cannot tell you who signs, whether there is budget this year, or what happens between a yes and a signature. High volume also hurts email deliverability, and in relationship-led markets like the Gulf a warm introduction still beats any sequence. MEDDIC still works
How should sales leaders use AI in their teams?
Start with where the week goes. Research, admin and reporting usually take most of it, so automate those first and keep qualification and forecasting with people. Judge success by selling hours and forecast accuracy rather than tool adoption. Buy tools whose value is data, and build narrow internal workflows where a model call does the job. Build or buy: the AI version of an old question
What is AI-first go-to-market?
Research, enrichment, call notes, drafting and reporting are handled by AI tools from day one, so salespeople spend more of their week selling and revenue grows without headcount following it. Qualification, judgement and relationships stay with people. AI enablement for revenue teams
How should a Western startup enter the UAE or Saudi market?
Treat the Gulf as a market, not a trip. Narrow down hard, because the UAE and Saudi Arabia buy differently, and so do Dubai and Abu Dhabi. Lead with warm introductions rather than cold outbound, put someone on the ground, design every pilot with an exit, and never count an MOU as revenue. What Western teams get wrong about Gulf go-to-market
Does cold outbound work in the UAE and Saudi Arabia?
It works far worse than in the US or UK. Buyers respond to warm introductions and relationships built over months, and enthusiastic meetings are often with people who cannot buy. Use outbound to open a door with a specific reason to talk, then ask early who else needs to be involved. What Western teams get wrong about Gulf go-to-market
How is B2B selling in the Gulf different from the UK or US?
Relationships carry more weight than process, a direct no is rare, and government and quasi-government buyers are large but slow and hard to forecast. Deals close quickly once the right person is involved. The UAE and Saudi Arabia buy differently, and so do Dubai and Abu Dhabi, so treat each as its own market. What Western teams get wrong about Gulf go-to-market
How do I hire a fractional head of sales or growth in Dubai?
Look for someone who has carried a number, built a sales motion from nothing and sold in the region, not only advised on it. A fractional lead usually works one or two days a week for three to six months, owns positioning, process and the early deals, and hands over a motion your team can run. Aarya Ravahi works this way as an embedded go-to-market lead.
When should a startup make its first sales hire?
When the founder has closed enough deals to write down who buys, why they buy, what they object to and how long it takes. That is usually somewhere between ten and twenty deals. The right first hire is often someone a few years in who has carried a number at a small company, not a VP from a large one. When to make your first sales hire
Does MEDDIC still work now that sales teams use AI?
Yes. AI tools solved research and admin, not whether a deal is real. Use AI for the research parts of MEDDIC, such as metrics, decision process and competition, and keep asking who signs and what has to be true for them to sign this quarter. MEDDIC still works
What go-to-market stack should an early-stage startup use?
A frontier AI model as the research layer, Attio or HubSpot as the CRM, Clay for building account lists from signals, a sequencer used with strict volume discipline, n8n or Make for automation, and call intelligence once someone other than the founder is selling. Skip intent data and annual-only contracts. The GTM stack I'd run at a two-person startup
How do you run go-to-market experiments to reach product-market fit?
Write down a hypothesis, a sample size and a decision rule before each test, then run many short tests, because most will fail and cycle time matters more than ideas. At Galadriel roughly thirty experiments ran over four months, and positioning changes beat channel changes. Thirty experiments to product-market fit
How do I find out why our sales have stalled?
Most stalls come from positioning, weak qualification, or a stack that creates activity rather than deals. A GTM diagnostic reviews your pipeline, call recordings, messaging and stack over one to two weeks and returns a fix list ranked by impact against effort. GTM diagnostic