The signals combine events from 20-21 October 2026 with a Eurostat dataset for 2025. Signed transactions are kept apart from closed ones, preliminary customs data from final results, and company forecasts from observed performance. The information does not constitute financial advice.
Signal of the day
On 20 October, Salesforce and Braze announced a definitive agreement under which Salesforce will acquire Braze for an enterprise value of about USD 5.66 billion. Braze shareholders would receive USD 47.50 in cash per share, 58.1% above the previous close. The transaction is signed, not closed: it requires a vote by Braze shareholders, regulatory approvals and the fulfilment of contractual conditions.
The traditional CRM keeps identity, history and opportunity. Braze operates closer to the behavioural moment: the product event, segmentation, decision and message. Value is shifting from simply storing data to orchestrating it. But integrating two platforms does not automatically produce a single customer view: taxonomies, identifiers, consent and commercial accountability must be harmonised.
For Romanian companies, the signal is not that a bigger stack must be bought. The question is whether the CRM, analytics, email, app, support and sales work from the same definition of the customer.
What Salesforce is actually buying
Companies have talked for years about a “single customer view”. In practice, data has stayed split between the CRM, the email platform, the app, analytics, support, e-commerce and sales. Each system sees part of the behaviour, and each department optimises its own indicator.
The stakes are moving towards the layer that can interpret an event and act: a customer abandons a process, returns to a product, changes frequency or signals a risk of leaving. This is the point at which marketing becomes commercial infrastructure. If identity, behaviour, permission, model and business rule are coherent, the company shortens the time between signal and response. If they are not, automation simply produces the same fragmentation faster.
The value of bringing functions together is not measured by how many of them move under the same brand, but by shorter integration time, fewer duplicates, a higher share of explainable decisions and the link between action and incremental revenue.
Data without clear rights is not an advantage
In an ecosystem of agencies, platforms, model providers, integrators and commercial partners, data rights become part of the economics of the offer. The contract must state who can use customer events, for what purpose, for how long and whether they can train or optimise an external model.
The same applies to IP: if a partner builds segments, rules, models or integrations, the company must know what remains portable at the end of the relationship. Otherwise, the initial speed becomes a switching cost.
Physical infrastructure sets the limit for software
NEXTDC sharply cut its FY2027 guidance after delays to grid works needed for a data centre in Melbourne: contracted revenue of about AUD 140 million has been deferred. Demand cannot be invoiced if the infrastructure is not available.
South Korea shows the other side of the cycle: in the first 20 days of October, semiconductor exports rose 222.5% year on year (preliminary customs data), and the daily average of exports by 65.9%. Capital is turning into real industrial demand, but concentration in a single category increases sensitivity to the AI investment cycle.
What scaling means in an ecosystem
An ecosystem is not a list of partners. It is a contract of complementarity: the company brings the process and the customer relationship, the technology provider brings the platform, the integrator shortens implementation time, marketing translates capability into a relevant problem, sales validates willingness to pay, and operations confirm that the promise can be delivered repeatedly.
Governance must distribute four things: decision, risk, data and value. If these remain ambiguous, the collaboration works in the pilot and breaks at scale.
Decisions for the next 30 days
- 1. Audit the data trail through to revenueData source, permission, identifier, decision, channel, customer response and financial outcome. KPIs: reconciled profiles, incremental conversion, retention, margin after cost to serve.
- 2. Treat energy and integration as commercial dependenciesFor any AI or automation project: firm connection date, redundancy, exit rights, cost of delay and risk owner.
- 3. Build partnerships with scaling thresholdsContract the pilot, data and IP rights, continuation criteria and unit economics. KPIs: time to first value, usage rate, contribution margin.