Personalization

This feature is currently rolling out in phases on Desktop, so not all users will see it yet. On iOS, it's in a closed beta with a limited group of testers. Some details may change before it becomes available to everyone.

Personalization helps Spark +AI features write and reply in a way that reflects your role, workflows, and how you usually handle email. Instead of producing generic text, Spark uses a personalization context about you: your role, your workflow, your scheduling preferences, and other relevant context to make AI-generated content feel more accurate and closer to what you usually do within the inbox.

What is Personalization

Personalization is a contextual snapshot of your role, workflow, communication patterns, and how you handle specific emails, which may significantly improve the quality of Auto-Drafts Spark creates for you. It covers things like your job, your time zone, how quickly you usually reply, how long your meetings tend to be, what projects you're working on, and anything else that helps Spark understand your day-to-day.

You can either:

Let Spark generate the context automatically by analyzing your recent emails
Provide your own context, manually, in plain language

Once your Personalization context is ready, Spark combines it with your My Writing Style snapshot every time a Spark +AI feature generates text for you. The two work together but do different things:

Personalization is about who you are: your role, preferences, and the rules you follow when handling email
My Writing Style is about how you write: your tone, greetings, sign-offs, and sentence structure

Personalization is currently used by the Auto-Drafts feature — you can learn more about this feature here. In the future, we plan to extend Personalization to other AI-powered features for more contextually relevant results.

Prerequisites

Before you can use Personalization, make sure:

You have Spark +AI enabled for your account (under Spark Settings)
You have enough email history in that account. If your mailbox is brand new or nearly empty, Spark won't have enough samples to generate a useful context, but you can still write your Personalization manually
You're running an up-to-date version of Spark on Desktop
Your Spark +AI quota is not depleted

Personalization is not enabled by default. You have to turn it on yourself when prompted by Spark, or by navigating to Settings > Spark +AI.

How to use Personalization

You can enable Personalization in two ways.

During onboarding. When you set up Auto-Drafts, Spark shows a Personalize Spark +AI step. Click Enable Personalization to turn the feature on. Spark will start building your context in the background and move you straight to the next onboarding step. If you'd rather skip this for now, click Skip and turn Personalization on later in Settings.
From Settings:

Open Settings > Spark +AI.
Select Personalization > switch the toggle on.

Spark will start analyzing your recent emails and show a loader that says "Personalizing your Spark +AI…" On mobile, you'll see a confirmation pop-up titled "Personalization is ready" once the context is generated.

Review your Personalization

When generation is finished, you'll see your Personalization context in Settings > Spark +AI > Personalization. Spark structures it around a few common areas:

Role: your job, location, timezone, and working hours. Example: “Engineer at ACME Corp, based in London, working hours 9–18 Mon–Fri”
How you handle emails: what you usually reply to, your typical length and tone, and the kinds of messages you deal with. Example: “Same-day to my team, skip cold outreach, 1–2 sentences, handle partnerships and updates”
How you handle scheduling: your default meeting length, your availability, and any scheduling link you share. Example: “30-min default, no mornings or weekends, respond with my scheduling link”
Other background: your industry, current projects, interests, or anything else worth knowing

How it stays current

Spark keeps what it learns about your role and habits in a single profile, so it doesn't relearn you each time, and uses that information to personalize AI features like Auto-Drafts. Based on your inbox, calendar, and Custom Instructions, it updates automatically every 30 days.

You can no longer edit the generated text directly. To correct or extend your Personalization context, add a Custom Instruction instead.

Add instructions to guide behavior, add facts, or set preferences: these override anything Spark learned automatically. You can edit or delete them anytime. For example:

Never suggest specific dates or times for meetings or calls in draft replies
Don't make decisions or commitments on my behalf when I'm asked for an opinion or a recommendation

Note: Instructions added via the Feedback Assistant chat (for example, when reporting an Auto-Draft) also appear here.

If you'd rather have Spark rebuild the whole profile based on your recent activity, see the "Regenerate from recent emails" section below.

Regenerate from recent emails

If your work has changed since you first enabled the Personalization feature — a new job, a new project, or a new way of running meetings — you can have Spark rebuild context based on your most recent emails. For this:

Open Settings > Spark +AI.
Select Personalization and tap or click Generate.

Note: Regenerating rebuilds the learned parts of your context from your recent emails and calendar. Your Custom Instructions stay in place and continue to apply: they're not affected by regeneration. Spark will show a confirmation alert before regenerating, so you can cancel if you'd prefer to keep your current context as-is.

Use case ideas

Personalization makes a difference across day-to-day email tasks. Here are a few situations where it really helps:

Sounding like yourself in Auto-Drafts. When Spark drafts a reply for you, it already relies on the analyzed context, so the draft comes out closer to what you would usually have written, with less editing on your part
Handling messages differently by sender type. If your snapshot indicates that you reply same-day to your team but skip other letters, Spark's suggested replies will reflect that priority. You spend less time deciding what gets a long reply and what gets a short one
Working across roles and projects. Tell Spark you're leading the project launch and consulting on a side project, and its drafts will refer to those projects accurately, using the right names, the right context, and the right level of formality
Onboarding into a new job. Update your Personalization context manually when your role changes, and Spark's drafts catch up immediately

Privacy

Here are the key points when it comes to privacy while using this feature:

No AI training. Personalization is not used to train AI models
No marketing use. Personalization is not used for marketing, advertising, or any purpose outside the Spark +AI features
Explicit consent. Personalization is off by default. It’s only active when you’ve explicitly enabled it.
Always reviewable and controllable. You can review the exact context Spark generated at any time, refine it by adding a Custom Instruction, regenerate it, or turn the feature off completely from Settings > Spark +AI > Personalization.

For more details, please review our Privacy Policy, which explains how your data is handled and protected.