diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
index 939e5341e..1d6357fb4 100644
--- a/CONTRIBUTING.md
+++ b/CONTRIBUTING.md
@@ -1,9 +1,13 @@
 # How to Contribute
 
-We'd love to accept your patches and contributions to this project. There are
-just a few small guidelines you need to follow.
+We'd love to accept your patches and contributions to this project.
 
-## Contributor License Agreement
+Notably, we're working to make this project contributor-friendly. Please follow the
+progress and comment [here](https://github.com/tensorflow/model-optimization/issues/131).
+
+## One-time Setup
+
+### Contributor License Agreement
 
 Contributions to this project must be accompanied by a Contributor License
 Agreement. You (or your employer) retain the copyright to your contribution;
@@ -15,6 +19,16 @@ You generally only need to submit a CLA once, so if you've already submitted one
 (even if it was for a different project), you probably don't need to do it
 again.
 
+## TensorFlow Model Optimization Specific
+
+### Contributing Whole Techniques
+
+See these [guidelines](CONTRIBUTING_TECHNIQUE.md).
+
+### Other Contributions
+
+Please wait until we finish this section.
+
 ## Code reviews
 
 All submissions, including submissions by project members, require review. We
@@ -25,4 +39,7 @@ information on using pull requests.
 ## Community Guidelines
 
 This project follows [Google's Open Source Community
-Guidelines](https://opensource.google.com/conduct/).
+Guidelines](https://opensource.google.com/conduct/). Please also
+look at the [TensorFlow contributor
+guide](https://www.tensorflow.org/community/contribute), in particular for the
+community values and style guides.
diff --git a/CONTRIBUTING_TECHNIQUE.md b/CONTRIBUTING_TECHNIQUE.md
new file mode 100644
index 000000000..6c9042815
--- /dev/null
+++ b/CONTRIBUTING_TECHNIQUE.md
@@ -0,0 +1,58 @@
+# Guidelines for Contributing a New Technique
+
+The following are guidelines for contributing a complete technique into the
+toolkit, for example, connection pruning.
+
+We will be evolving these guidelines to make the process more effective, and
+as we receive more contributions. For example, we are working on creating a
+repository of training scripts for
+different models and tasks aimed at simplifying technique validation
+([issue](https://github.com/tensorflow/model-optimization/issues/133)),
+making the project contributor-friendly
+([issue](https://github.com/tensorflow/model-optimization/issues/131)), and
+having reproducible results.
+
+
+## Contribution Components
+
+1. Please start by providing an [RFC](https://github.com/tensorflow/community/blob/master/governance/TF-RFCs.md) under [model-optimization/community/rfcs](https://github.com/tensorflow/model-optimization/blob/master/community/rfcs).
+   Consider the following guidelines:
+   * API and implementation should strive for similarity with existing
+     techniques to provide the best user experience.
+   * consider the end-to-end experience for the user of your technique.
+   * be prepared for a potential design discussion.
+
+2. Provide experimental results that demonstrate benefits to end-users across
+   models and tasks. This is probably the main criteria for us to consider, so
+   the stronger the validation the better. Some relevant aspects are:
+   * for Keras APIs, we recommend the following test tasks (and
+     hope to be adding more):
+     * [BERT task](https://github.com/tensorflow/models/tree/master/official/nlp/bert)
+     * [object detection](https://github.com/tensorflow/models/tree/master/research/object_detection)
+   * results in combination with other techniques (e.g. post-training integer
+     quantization).
+   * results include not only accuracy but also deployment metrics (e.g. model,
+     storage space, latency, memory, to mention a few).
+   * reproducible results are best: e.g. provide hyperparameters with minimal
+     scripts to reproduce results.
+   * when possible, include trained models that showcase those benefits.
+
+3. Documentation and tutorials:
+   * overview page that requires minimal end-user domain knowledge. [Sample](https://www.tensorflow.org/model_optimization/guide/pruning)
+     * TODO(tfmot): template
+   * colab tutorial that covers the most common use cases and user
+     journeys. [Sample](https://www.tensorflow.org/model_optimization/guide/pruning/pruning_with_keras)
+   * advanced documentation that may cover:
+       * advanced use cases not in tutorial. [Sample](https://www.tensorflow.org/model_optimization/guide/pruning/train_sparse_models)
+       * internals not relevant to end-user (e.g. app and model developers) but relevant to
+         others in ecosystem (e.g. hardware developers and other contributors).
+
+4. Packaging and release:
+   * releases are managed by the TensorFlow Model Optimization team. Work with
+     them to produce releases.
+   * auto-generated API docs.
+
+5. Collaborative blog post (optional)
+   * samples: [pruning
+     API](https://medium.com/tensorflow/tensorflow-model-optimization-toolkit-pruning-api-42cac9157a6a)
+     and [post-training integer quantization](https://medium.com/tensorflow/tensorflow-model-optimization-toolkit-post-training-integer-quantization-b4964a1ea9ba)
diff --git a/google3/third_party/tensorflow_model_optimization/opensource/community/rfcs/README.md b/google3/third_party/tensorflow_model_optimization/opensource/community/rfcs/README.md
new file mode 100644
index 000000000..b069eda98
--- /dev/null
+++ b/google3/third_party/tensorflow_model_optimization/opensource/community/rfcs/README.md
@@ -0,0 +1,7 @@
+# TensorFlow Model Optimization RFCs
+
+This directory stores approved RFCs for new techniques.
+
+## Process
+
+Please carefully read our [new technique contribution guidelines](https://github.com/tensorflow/model-optimization/blob/master/CONTRIBUTING_TECHNIQUE.md).