A simple way of thinking about this is that a sample of a monoclonal antibody comprises millions of identical antibody molecules, whereas a polyclonal antibody includes many antibody molecules with different primary sequences, though still all recognizing a single antigen. Both are used for conjugation. There are pros and cons. Monoclonals allow more consistent performance over long time periods (since different batches in theory are the same); polyclonals may show more variation across batches, even if the conjugation method is constant.

An advantage of polyclonals is that you have a mixture of antibodies with different attributes (affinities, kinetics etc.), and thus different subpopulations may work well in different applications. By contrast, the molecules of a monoclonal Ab are identical, and if your application requires, for example, a fast on rate, a monoclonal with a slow on rate won't be suitable at all. By contrast, a polyclonal will have a range of on rates and affinities, and possibly a broader range of application.

When attaching labels to either type of antibody, you will have many different CONJUGATE molecules, potentially with varying attributes (affinity, even selectivity). This is true even with mAbs (despite the common primary sequence) since the labels will be attached at different sites on individual antibodies molecules (unless strategies are used to achieve a definite point of attachment - which is not commonly done). Thus, your next batch of conjugate will always be slightly different, especially if the degree of conjugation is not carefully controlled.

The principle worry with polyclonals is the certainty that a good batch WILL run out at some point. A clever way round this issue is to create defined polyclonals by mixing together several monoclonal Abs in fixed proportions.

Finally, secondary antibodies will generally be polyclonals; primary antibodies could be either type. Both types of Abs can be great or awful before and/or after conjugation.