Adding and Configuring Content Steps in AI Studio

If you’re using one of Autobound’s pre-built campaign templates (for example, Multi-Step Email Sequence or Single Personalized Email), a content column is automatically created for you.
​

If you’re starting from scratch or want to add additional messaging steps, you can do that using the Actions button inside your campaign table.

To create a new content step:

Once added, a new column will appear in your table.

Before content can be generated, Autobound needs to know which fields to use for both the contact and (optionally) the seller.
​

At minimum, you must map at least one of the following contact identifiers:
​

• Contact Email

• Contact LinkedIn URL

• Company Domain

Ideally, include all three for the best resolution accuracy.
​

To map a field:

Repeat this process for each input you want to map.
​

If you want to generate content on behalf of a specific seller, you can also map Seller Email, Seller LinkedIn, or Seller Company URL. If you’re generating content as yourself or don’t need that distinction, you can leave these blank.

Alternatively, you can map the seller name in the Name Accuracy section — this is only helpful if you want to include a sign-off in the email such as “Best, NAME.”

The Content Inputs section is where you define any additional context that should influence the messaging.

The most important field here is Additional Context.

This is your “last mile” input — it can include data or insights pulled from your Custom Researcher, your CRM, or any other column that adds helpful context to the message.

Autobound will automatically reference this input when generating personalized content.

Other optional inputs include Value Proposition, Sales Asset, Writing Style, and Language. It’s recommended not to touch these unless you have a very good reason to override Autobound’s default logic, which automatically pulls from your Content Hub to determine the most relevant information to use.

By default, Autobound matches the persona, industry, and use case of each contact using your Content Hub data.
​

If you choose to map your own columns to these fields, they will override the defaults. In most cases, it’s best to leave them blank unless you have a specific reason to do so.

Next, you’ll see a toggle to enable Content Hub Preferences.

Turning this on opens a configuration modal where you can:

• View your current Content Hub defaults.
​

• Select specific Value Props, Sales Assets, Writing Styles, or Languages to apply across the entire campaign.
​

• Add campaign-level context in the Know Something We Don’t? field (this will be overwritten by data passed to Additional Context if you leverage that input at the row level in your campaign).

This helps Autobound tailor every message to the campaign’s purpose.
​

For more on customizing value props and assets, see: Managing Your Content Hub

Once your inputs are configured, you can add additional message steps for your sequence.
​

Each step can be customized individually:
​

• For email steps, choose New Thread or Reply.
​

• Click Add or Edit Blueprint to define a custom prompt or guidance for that step.
​

Blueprints control the tone, structure, and goal of each message.
​

AI Blueprints are entirely optional. If you do not create a blueprint, Autobound will still generate a fully personalized message based on your Content Hub, insights, research, writing style settings (if selected), and any additional context. The output will still be high quality without a blueprint.

Use a blueprint if you want more fine tuned control over how a specific step should look. There are two ways to use blueprints depending on the level of structure you want:

If you have specific language you want the message to resemble, paste your template into the blueprint. Autobound will keep roughly 85 percent of your original text and weave in relevant insights, personalization, and step context.

Use square brackets (for example, [insert something personalized here], [ask a question about the workflow challenge]) to tell Autobound where it should insert contextual details.

If you prefer to tell the LLM what to do rather than provide a template, you can write instructions inside square brackets. Autobound will follow your guidance while still generating original copy.

This gives you more creativity and flexibility for steps where tone and purpose matter more than an exact template.

Below are three example blueprint prompts you can use as a reference.

EMAIL 1 – Personalized Intro and Value Alignment

[Write a short, natural subject line (4 to 5 words max) that connects directly to why you are reaching out right now. It should sound like something a peer would actually say. Pull from research. If there is no strong research, use a grounded observation. Avoid jargon or buzzwords.]

[instructions: write a concise, conversational cold email under 75 words. Use short, clear sentences and distinct line breaks. Plain, natural language only.]

Hi {{prospectName}},

[first paragraph: reference recent, highly relevant research or information about the prospect or their company that ties directly to their current priorities. Be specific. If no meaningful research is found, open with a credible observation about a common friction in their industry.]

[second paragraph: ask one neutral, curiosity driven question linked to a real workflow or performance challenge your product solves. Avoid pushiness.]

[third paragraph: describe how your product helps solve that problem, focusing on outcomes. End with a soft CTA like “Worth exploring?” and an optional out.]

Best,

{{userName}}

EMAIL 2 – Case Study and Social Proof

[instructions: write a conversational follow up email under 70 words. Use different context than email 1.]

Hi {{prospectName}},

[start with one short line referencing a relevant customer story. Include company name and measurable result if available.]

[explain briefly how that customer achieved the outcome using your product. Connect it to the prospect’s world. Use different research than in email 1.]

[if no case study is available, use general credibility such as “teams in your space” but never invent details.]

[end with a low friction CTA.]

Best,

{{userName}}

EMAIL 3 – Friendly Bump

[instructions: write a short, friendly bump email under 40 words. One paragraph only. Casual tone.]

{{prospectName}} - [acknowledge previous outreach. “Just wanted to circle back in case this slipped through.”]

[mention one concise, value driven reason your product may still be relevant. Use fresh information or research not used before.]

[end with a soft yes or no question.]

-{{userName}}

You can use Run Conditions to control when each content step executes.
​

For example:
Use AI to write your formula: “Only run this step if the contact's email: {{Contacts email}} does not contain gmail.com.”

This is ideal for filtering high-value prospects or selectively running actions based on data in your table.

Autobound automatically chains content steps together. Each step has context on the messages before and after it, ensuring that your sequence reads naturally and consistently throughout.
​

This means you don’t need to manually toggle between sequence or individual modes — the system intelligently manages context for you.