New Meta pixel with no purchase data — best optimization strategy for a new store?

Hi everyone,

I recently launched my Shopify store selling digital PDF guides (~$9 CAD each) and set up my first Meta ad campaign. My pixel is brand new, so it has no purchase history yet.

Because of this, I couldn’t optimize for Purchases — Meta had nothing to learn from — so I switched my ad set to optimize for Landing Page Views instead. My first small test ($5/day) delivered landing page views at around $0.20 each, but no purchases yet.

My questions:

  1. Is optimizing for Landing Page Views the right move for a new pixel, or should I optimize for Add to Cart / Initiate Checkout as an intermediate step?
  2. How many purchase events does the pixel need before switching the campaign to Purchase optimization makes sense? I’ve seen “50 conversions per week” mentioned — is that per ad set?
  3. Is it better to just optimize for Purchases from day one and let the pixel learn slowly, even with no data?
  4. Any tips specific to low-priced digital products where the whole funnel is one page?

My pixel is installed via the official Facebook & Instagram app on Shopify and events (PageView, ViewContent) are firing correctly — verified in Events Manager.

Thanks in advance!

I’d optimize for Purchase from day one. Landing Page Views will find visitors, while Add to Cart or Checkout optimization may find people who take those actions without buying.

The 50 conversions figure is per ad set per week, but it’s a learning-phase benchmark—not something you need before selecting Purchase. At $5/day, keep everything consolidated into one campaign and one ad set.

Also complete a test order to confirm the Purchase event, value and currency are firing correctly. For a $9 CAD product, the bigger profitability lever may be increasing order value with a relevant bundle rather than only trying to improve pixel learning.

You’re on the right track, but I wouldn’t optimize for Landing Page Views for too long. That objective mainly teaches Meta to find people who click and load pages, not necessarily people who buy.

For a $9 digital product, I would usually do one of these.

• If your budget can support it, optimize for Purchases from the beginning. Meta’s algorithm has improved a lot and can still find buyers even with a fresh pixel, although learning will take longer.

• If your budget is very limited and you’re getting almost no conversions, optimizing for Initiate Checkout can be a reasonable temporary step. I generally skip Add to Cart because it doesn’t always correlate with completed purchases.

The “50 conversions per week” recommendation is a guideline for the optimization event at the ad set level. It isn’t a strict requirement. Plenty of smaller stores still optimize for Purchases without reaching that number every week.

Since your funnel is only one page, I’d also spend time improving the page itself. With low priced digital products, the biggest gains often come from increasing the conversion rate rather than getting cheaper clicks. Your headline, value proposition, trust signals, testimonials, preview images, and checkout flow all have a big impact.

Out of curiosity, what’s your current landing page conversion rate, and how many landing page views have you received so far? That would make it easier to tell whether this is mainly a traffic issue or a page conversion issue.

Landing page views: ~25-30 total (last Ads Manager reading showed 25 impressions; landing page views were showing “—” so likely under 25)

Sales: 0

Conversion rate: 0% (0 sales ÷ ~25 views

I wouldn’t use Landing Page Views for long. It mostly teaches Meta to find cheap clickers, not buyers.

With ~25 visits you don’t really have a conversion problem yet, you have no sample size. I’d first place a real test order and make sure Purchase, value, currency, and thank-you-page tracking are perfect.

Then run one clean Purchase campaign/ad set and judge the landing page after a few hundred real sessions. For a $9 PDF, the bigger lever is probably bundle/AOV and showing a preview of the guide, not trying to warm the pixel with weak events.