Verke Editorial

Social anxiety exercises you can practice on your own

Verke Editorial · May 5, 2025

This is an 8-week self-guided program. The exercises are drawn from the same CBT protocols therapists use — specifically the Clark & Wells model that targets the cycle keeping social anxiety alive. They work in a specific order: first you learn to see the pattern, then you start testing it, then you learn to debrief without spiraling. Skipping ahead is tempting but counterproductive — the cognitive prep in weeks 1–2 is what makes the exposure in weeks 3–8 productive rather than just painful.

Exposure without structure is just suffering. Structure without exposure is just planning. You need both, in the right sequence.

You'll need a notebook (physical or digital) and 15–20 minutes, 3–5 times per week. That's it. If you want the theory behind why these exercises work, start with what social anxiety is and what actually helps. This page is the practical toolkit.

Before you change anything, you need to see what's happening. Your brain is making predictions about social situations — most of them automatic and most of them wrong. These two exercises make the invisible visible.

Build your exposure ladder (session 1)

This is the backbone of the whole program. An exposure ladder ranks social situations from least to most anxiety-provoking so you can work through them systematically instead of jumping into the deep end.

Step 1: list 10–15 social situations that trigger anxiety for you. Step 2: rate each one 0–100 on the SUDS scale (Subjective Units of Distress — 0 means no anxiety, 100 means the worst you can imagine). Step 3: order them from lowest to highest. Step 4: find your starting point — the first rung that hits 30–40 SUDS. Below that is too easy to teach your nervous system anything new. Above it and you risk flooding rather than learning.

A worked example: ordering coffee while making eye contact (15) → asking a colleague a question (25) → calling a friend instead of texting (35) → attending a small gathering (50) → starting a conversation with a stranger (60) → giving a toast at a dinner (80). Your numbers will be different — what matters is that the ladder is yours. If you're looking for a dating-specific ladder, see dating with social anxiety. For workplace scenarios, see performance anxiety.

Start a thought record (sessions 2–4)

A thought record captures the automatic predictions your brain makes in social situations. Use a five-column format: Situation → Automatic thought → Emotion plus intensity (0–100) → Evidence for and against → Balanced alternative. The goal isn't positive thinking — it's realistic thinking.

Worked example: "Meeting new people at a party" → "I'll stand there with nothing to say and everyone will notice" → Anxiety 75 → For: "I've had awkward silences before" / Against: "Last time I talked to two people and it was fine" → "I might have some quiet moments, and I'll probably also have some conversations."

Do 3–5 thought records this week. You don't need to be in the situation while you write them — do them from memory after a social interaction, or before one you're anticipating. The goal right now is to start noticing the predictions your brain makes automatically, not to fix them.

Now you start doing the things — but structured, not white-knuckled. Each exposure is a behavioral experiment, not an ordeal. You're not trying to survive it. You're trying to find out whether your prediction was accurate.

The prediction test (every exposure, starting now)

Before each exposure, write: "I predict [specific outcome]." Rate your belief in that prediction 0–100. Then do the exposure. Afterwards, write: "What actually happened was..." Rate your belief in the original prediction again 0–100.

Keep these records. The stack becomes your evidence base — physical proof that the predicted catastrophe rarely matches reality. Start at rung 1–2 of your ladder. Aim for 3–5 exposures per week. Research recommends this frequency because regular practice prevents anxiety from rebounding between sessions.

The safety behavior audit (session 5 or 6)

Safety behaviors are the subtle things you do to manage anxiety in social situations. They feel like coping. They're actually what prevents the belief-disconfirmation that makes exposure work. Common ones: avoiding eye contact, speaking quietly, keeping your phone out as an escape route, only attending events with a "safe person," having drinks before social situations, or rehearsing conversation topics obsessively.

List your top five. Then pick one — just one — to consciously drop in your next exposure. Observe what happens. The experiment: does the predicted disaster occur without the safety behavior? The counterintuitive finding from the research is that dropping safety behaviors usually reduces anxiety, not increases it. For why this works, see the maintenance cycle.

You're building on evidence now, not starting from scratch. The prediction tests from weeks 3–4 gave you data. Now you use that data to tackle progressively harder situations while adding two more techniques.

The attention shift (use during every exposure)

Social anxiety locks your attention inward — monitoring how you look, how you sound, whether your hands are shaking. The attention shift breaks that loop. In any social situation, redirect your focus outward: notice three things about the other person. What are they wearing? What's their expression? What did they just say?

This isn't mindfulness advice — it's a specific CBT technique from the Clark & Wells model. The anxious brain can't self-monitor and genuinely engage at the same time. Each outward observation is a micro-experiment that weakens the self-focused attention link in the maintenance cycle. Start in low-stakes settings — with a barista, a familiar colleague — and move to higher-stakes situations as you climb the ladder. For the workplace-specific version, see performance anxiety.

Drop one safety behavior per week

Work through your audit list from weeks 3–4. Week 5: drop behavior number one. Week 6: drop number two. By week 8, you've tested three or four of them — and you have evidence for what actually happens without the crutch. Wells and colleagues found that social interactions without safety behaviors produced lower anxiety ratings and more positive self-appraisals than interactions with them. The safety behavior feels protective. The data says otherwise.

Move up 1–2 rungs per week

If a rung no longer triggers 30+ SUDS after two or three exposures, it's done. Move up. If you're stuck, check: are you using safety behaviors that blunt the learning? Is the rung actually too high — SUDS 70 or above? If so, insert an intermediate step. The sweet spot for learning is 30–50 SUDS — enough challenge to activate new learning, not so much that the nervous system floods and shuts down.

This isn't a phase — it's the constant. Do this after every exposure, all eight weeks. The structured debrief replaces the biased post-event replay that social anxiety runs automatically. Left unchecked, that replay selects the worst ten seconds of an interaction and loops them until they feel like the whole experience. The debrief interrupts the loop with structure and a time limit.

The protocol

Five questions, maximum five minutes — set a timer. (a) What did I predict? (b) What actually happened? (c) What went better than expected? (d) What would I do differently next time? (e) Close the notebook.

Beyond five minutes is rumination, not reflection. The time limit is the whole point. Unstructured post-event processing is how social anxiety maintains itself — Penney and Abbott's 2014 review found large effect sizes for structured interventions that target exactly this phase. If the thoughts keep looping after you close the notebook, shift to a physical activity — walk, exercise, cook. Give the nervous system something else to process. For more on breaking the replay loop, see replaying conversations in your head.

Your progress log

Keep a simple table: date, situation, prediction belief before (0–100), prediction belief after (0–100). Over four to eight weeks, the "before" numbers trend down. That's not wishful thinking — that's your threat model updating based on evidence you collected yourself. The log makes the trend visible, which matters on the days when progress feels invisible.

Self-guided work is effective for many people — but it has limits. Clear signals that it's time to add professional support: SUDS scores aren't decreasing after six to eight consistent weeks. Panic attacks during or before exposures. Avoidance is increasing despite the program. Substance use to manage the anxiety. Depression making it hard to start.

Self-guided practice plus therapy is more effective than either alone. Adding help isn't failure — it's optimizing the approach. For more on professional help thresholds, see when to seek professional help.

If you want a practice partner for the cognitive prep, the exposure planning, and especially the post-event debrief, Judith is built for this. She uses the same CBT framework this program is drawn from — and she's available at 11 p.m. when the post-event replay starts running. She remembers what you've been working on across sessions, so the work compounds. For more on the method, see Cognitive Behavioral Therapy.

Chat with Judith about this — no account needed