Digital programs - no therapist, no weekly check-ins, no group sessions - are what most people actually reach for when they want to do something about their drinking. Alcobalance is one of them: a set of tracking tools and behavior techniques you work through on your own. The appeal is obvious: available whenever, no waitlist, no appointment. But what exactly should go inside one of these? I think about that question a lot. A team of researchers just answered it more rigorously than I could.

What they did

Researchers assembled a panel of 14 experts in behavioral science, alcohol and tobacco treatment, and digital interventions. They ran two rounds of the Delphi method (a structured consensus process): panelists rate independently in round one, see the group results in round two, and can revise their positions.

The pool was 20 behavior-change techniques (BCTs). Each was rated against the APEASE criteria - acceptability, practicability, effectiveness, affordability, safety, and equity. The consensus bar was 70%: at least 10 of 14 panelists had to agree on every criterion, not just overall.

One constraint shaped the whole thing: the target was a one-time, self-guided digital intervention. No therapist in the loop, no follow-up sessions scheduled.

What they found

Six BCTs cleared every bar:

  • Goal setting. A specific reduction target - not "I want to drink less" but "no more than two drinks on Friday."
  • Individual plan. A personal map: when, how, and what to do instead of the usual habit.
  • Reduction strategies. Concrete tactics for specific situations - what to do at a party, what to do on Friday after work.
  • Feedback. Your numbers compared against your own goal - not a population average, your data against your plan.
  • Reattribution. Reconsidering what actually triggers the habit: "I drink to unwind" and "I drink automatically when I'm anxious" are two different levers.
  • Pros and cons. An explicit analysis of the case for and against changing - structured weighing, not a background feeling of "maybe I should."

Several BCTs reached partial consensus - cleared some APEASE criteria but not all. Eight of the 20 were flagged as unsuitable for a one-time, unsupervised format: some require repeated sessions, some need a live clinician.

What it means

Reattribution is the most non-trivial of the six. People often can't name what actually triggers their habit - is it anxiety, boredom, or just a reflex tied to a specific situation? Once the trigger is named, there's a concrete lever. Without that step, any action plan floats: you know what to do but not when or why.

The two planning techniques are split intentionally. The individual plan sets the overall strategy; reduction strategies fill in the situational specifics - "if the group orders another round, I do X." Without that level of detail, a plan stays an intention.

All six share a thread: specificity. The feedback step works because it measures you against YOUR goal, not a group average. The plan and tactics are yours - built around your habit, not a default.

That's the logic behind Alcobalance: give you the tools to tune the approach to your own habit, show you the full picture, and let you steer from there.

Source: JMIR Formative Research, DOI