How Do We Connect an Ecosystem?

Nine months facilitating the DemocráTICa Community and working with its Network of Specialists left me with an uncomfortable conclusion: almost everything the sector does to connect an ecosystem operates at the level where the least can be moved.

by Mariel Morales Lezica

When an organization sets out to strengthen an ecosystem, the first reflex is almost always the same: do a mapping. A survey of who does what, where, with what capabilities. I've done mappings. I still stand by them. But over the past nine months, leading the Community component of the DemocráTICa Network, I learned that mapping solves a problem that, often, isn't the actual problem.

The problem wasn't that they didn't know each other

The program's regional diagnosis found what we called islands of excellence: specialized capabilities — in artificial intelligence, in global governance, in territorial methodologies — distributed among organizations that, for the most part, don't know one another. Up to that point, mapping is the right answer.

But the Strategic Mapping of the Network of Specialists, based on an institutional survey and in-depth interviews with six reference organizations, returned a more demanding finding: high strategic complementarity and low structural overlap. Translated: the ecosystem's limit isn't a lack of capacity. It's the absence of connective tissue between capacities that already exist.

And where organizations did know each other, a second condition showed up, which we named relational latency: mutually recognized shared agendas that never activated into joint work.

The case that shows this best is Fundación Corona and HIP. Both had been part of the same Core Team since the program began. It wasn't a lack of information about each other. It wasn't a lack of capacity. They hadn't identified a point of joint work until the mapping made the complementarity explicit — Corona's results-based financing mechanisms applied through territorial methodologies, and HIP's donor networks — and from that recommendation they started developing a joint initiative.

What was missing wasn't knowing each other. It was someone naming the complementarity out loud, and there being an occasion to do something with it.

The methodological consequence

If the obstacle is latency and not unfamiliarity, the deliverable changes. It isn't a directory: it's a facilitated, recurring occasion where a latent relationship has to become a concrete request, offer, or commitment.

Every design decision for the six Community gatherings came from there. Each one was built around a written script with timed blocks, a named facilitation technique, an assigned facilitator, and a harvest column filled in afterward. We held to four commitments across all six:

• Conversation before presentation. In none of the six gatherings did the presentation take up most of the time.

• A replicable social technology each time. Alliance Fair, Troika Consulting, Design for Wise Action, meaning rooms, dialogue tables, live collective mapping. Each method was chosen with an explicit criterion: that a participating organization could use it again on its own, without us.

• Collective harvest as a closing block, not as minutes written up afterward.

No extraction. The rule, written that way in our internal documents, was that every gathering had to define what it gives to the organizations, "so as not to fall into an extractive logic of getting people together just because."

Read in sequence, the six designs escalate what they ask of participants: from leaving an offer on a shared wall, to advising a peer, to spending an hour improving another organization's project, to mapping how the Community protects itself.

But the most concrete work didn't happen in the gatherings. It happened between them: reading organizations' requests and offers as a matching problem and sending targeted introductions — what we called threads of the fabric. Three documented rounds, on financing, technological development, and strengthening leadership.

Two findings that changed our approach

One: the method of convening explains participation far more than the size of the support.

We learned this firsthand. The first session of the Network of Specialists had lower attendance than expected. We logged it for what it was: we hadn't made the space's value proposition explicit enough. We redesigned it with a shorter framing and one-on-one conversations with each organization beforehand, and convened it again. Five of the six invited organizations took part, and the session produced the first structured record of what each could contribute to the ecosystem and what it sought from it.

Three of the Specialists Fund's initiatives reached the same conclusion independently, without talking to each other: closed, trust-based convening outperformed open convening. In one case, 20 of the 25 registered organizations actually took part in a closed strategic dialogue. With grants of between $2,500 and $4,000.

Two: an open community doesn't self-organize just because it's open.

Without structure, it risks acting like a sponge: absorbing participation without converting it into exchange. That finding explains the shift we made over the period, from open convening toward more curated, relationship-based forms.

Activating beats creating

The Specialists Fund supported seven collaborative initiatives with $21,500 total. Six of the seven worked with a network, movement, or community that already existed. Only one built a genuinely new network — 127 young people across 18 of Peru's 24 regions, ahead of the elections.

That figure seems to me the most important in the whole portfolio, because it runs against funding's bias toward the new. The Fund's characteristic contribution wasn't creating collaboration: it was activating latent collaboration. And what the money paid for wasn't infrastructure or technology. It paid authorship fees, facilitation, coordination time, stipends. It paid the transaction cost of coordinating — exactly what almost no funding instrument covers.

Where the leverage is

It helped us to read the problem as an iceberg: events, patterns, structures, and mental models. Events are the most visible level and the least leveraged. Convening more gatherings is the weakest lever there is.

The strong levers operate further down. At the level of structures: allocating resources specifically to coordination, not to organizations separately; installing a role that absorbs the transaction cost of coordinating; circulating information about capabilities and complementarities. And at the level of mental models, where collaborating stops reading as a cost and starts reading as an investment — the shift from an ego-system logic, organized around competition for limited resources, to an eco-system logic, organized around the resilience of the regional fabric.

This is also a critique of our own measurement framework. As long as we measure and fund organization by organization, fragmentation remains structurally invisible. And therefore unresolved. The reductionist lens — analyzing, measuring, and funding each organization separately — is essential for accountability. It isn't enough to understand an ecosystem.

What we still haven't achieved

It would be easy to end here. It wouldn't be honest.

Our own sessions exposed the main limit: information flows are still concentrated in the program team, and coordination between member organizations still depends on our facilitation as a bridge and catalyst. The group itself identified this as a problem — not us.

Across the portfolio of seven initiatives, there's only one documented connection between them, with no follow-up joint work: the portfolio didn't become a portfolio. And we have no documented public-policy change. Nine months isn't enough for that, and saying otherwise would confuse activity with impact.

That limit defines the next stage of the methodological work rather than closing it. Because the proof that facilitation is working isn't that people keep needing it. It's the opposite. That's why the Community Directory — which the member organizations themselves requested — was designed explicitly so that "the Community itself can self-manage collaborations without depending on central coordination." In other words: designed to make our own facilitation progressively unnecessary.

So, how do we connect an ecosystem?

Not by convening it. An ecosystem doesn't get connected with more gatherings, more mappings, or more open calls.

It gets connected by lowering the cost of collaborating: naming the complementarity that's already there, paying for the coordination time nobody wants to fund, and creating occasions where a latent relationship has to turn into a concrete commitment.

The fabric isn't built. It's activated.