jevmarketing

An open-source marketing toolkit

Seven marketing
decisions.
One small engine.

Find the pattern in your ads, briefs, search terms, and leads. Turn supplied data into decisions you can inspect.

Explore the workflows Read the source

Python toolkit · Optional live Jev scoring · MIT licensed

THE DECISION LAYER7 workflows / 1 toolkit
YOUR INPUTJEV + CODEYOUR REVIEW
Ad tags Longevity Brief scoring
j.JEVdecision engine
Search terms Creative fatigue Landing match Lead fit
Structured input. Explicit criteria. Reviewable output.

Your data, in.Imported JSON or CSV.

The reasoning, visible.Evidence beside each decision.

The next move, yours.Review before any account changes.

Inside the toolkit

From input to a clearer next step.

Synthetic examples

Choose a workflow. Inspect the source records, read the evidence, and take the JSON with you.

Imported creative records

Ad library tags

Group supplied ads by their hook, format, and offer.

Deterministic demo, no model call. Frozen Python results, synthetic records. Demo confidence is illustrative, so model-based examples route to review.

Loading the saved examples…

KEEP IN MIND

This browser only reads saved examples. Run the CLI with your own records for a new analysis.

A small, deliberate division of labor

Let code count.
Let Jev classify.
Keep the evidence.

Use ordinary calculations for dates, denominators, and CTR. Add a decision model where interpreting language matters.

See how it works
01

Bring a bounded question

Supply the records and the context: your brand, customer profile, or the landing-page text you want compared.

02

Run a focused workflow

The demo uses reproducible rules. Live mode uses Jev for supported classification and scoring, alongside transparent calculations.

03

Review the decision

Inspect the evidence and limitations. Low-confidence results need a human. The toolkit does not write to your ad accounts.

Run it yourself

Start with the sample.
Then bring your own.

Python 3.11+. Run the demo without an API key. Add your Jev credentials when you are ready for live classification.

Live configuration
TERMINAL
git clone https://github.com/stas4000/jev-marketing.git
cd jev-marketing
python3 -m venv .venv
source .venv/bin/activate
python -m pip install -e .
python -m jev_marketing demo

No API key needed for the demo.

Source & scope

Useful decisions start
with honest boundaries.

Inspired by this seven-workflow marketing post. Built as an independent, open-source implementation by Stas Sorokin.

The examples demonstrate behavior, not business performance. Longevity is not profitability. A fit score is not a conversion probability. A negative keyword is a candidate for review.

No ad-library scraping, account writes, conversion uploads, or performance guarantees. Live mode sends supplied context to your configured Jev provider. This static demo has no form, account connection, or analytics.

Read the limitations ↗Jev documentation ↗